MANAGEMENT OF
UNCONTROLLED HAZARDOUS
      WASTE SITES

-------
       THE 5TH NATIONAL CONFERENCE ON
        MANAGEMENT OF
UNCONTROLLED HAZARDOUS
            WASTE SITES
         NOVEMBER 7-9, 1984 • WASHINGTON, DC
                    AFFILIATES
         U.S. Environmental Protection    American Society of Civil Engineers
         A9ency          Association of State & Territorial
         Hazardous Materials Control      Solid Waste Management Officials
         Research Institute        National Association of Local
         U.S. Corps of Engineers       Governments on Hazardous
         U.S. Geological Survey       Wastes
         Federal Emergency Management   Portland Cement Association
         Agency          Centers for Disease Control

-------
                                    PREFACE
  CERCLA will place the states in the implementing role and will delegate responsibilities to
the EPA Regional Administrators. In the implementation of the CERCLA programs, new sites
will be identified and new technologies will be developed and employed.
  Under CERCLA, the U.S. EPA has three major elements of its strategy. First, uncontrolled
hazardous waste sites in the Agency's current inventory will be assessed. Second, those sites
which present an imminent threat to public health or the environment will be stabilized. Third,
those sites that should receive priority attention for remedial clean-up action will be dealt with
first, using the National Contingency Plan for guidance.
  As of July 1984, EPA indicated that they had begun remedial investigations and feasibility
studies at 258 NPL sites, 60 NPL sites are being cleaned up, and emergency clean-ups had been
initiated at 400 sites, NPL and non-NPL. In addition, seven sites on the NPL are being cleaned
up  by private groups under  EPA supervision.
  In FY 1985, CERCLA is expected to be extended at an increased funding level. Much of the
increase in these resources will be devoted to expansion of remedial construction projects at
NPL sites. During FY 1985, EPA plans to begin construction work at 46 sites, compared with
15 sites during FY  1984. By the end of FY 1985, EPA expects to complete or have under way
planning activities  for cleanup at 396 NPL sites.
  In October 1981, EPA published an interim priority list of 115 sites and in July 1982, ex-
panded the eligibility list adding 45 sites for a total of 160 sites. EPA published a list of 418 sites
as a proposed rule in December 1982, including 153 of the 160 sites previously published. Times
Beach, Missouri, was proposed in March 1983, bringing the total proposed to 419. After a
period of public comment, EPA published the NPL as a final rule in August 1983. At the same
time, EPA proposed 133 new sites in its first NPL update. In the second NPL update, released
October 1984, EPA proposed 128 sites be added to the NPL and modified some  existing sites.
The NPL, as of November 1984, consists of 538 confirmed and 244 newly proposed sites plus
four remaining from the 1983 proposed list. This second update is included in the Appendix.
Refer to the 1983 Proceedings of the National Conference on Management of  Uncontrolled
Hazardous Waste Sites for the original NPL and the October 1983 update.
  The papers presented at the National Conference on Management of Uncontrolled Hazar-
dous Waste Sites update the significant technology and information necessary to identify and
evaluate uncontrolled hazardous waste sites and control and mitigate the consequences from
those sites on the National Priorities List. These Proceedings emphasize actual experience ob-
tained  during the various stages necessary for remediation of the numerous SUPERFUND
sites.

-------
                       ACKNOWLEDGEMENT
  We would like to express a hearty "thank you" to all the individuals and organizations who
assisted in developing the program, the proceedings and the success of the 5th National Con-
ference and Exhibition on Management of Uncontrolled Hazardous Waste Sites.
  Affiliated Organizations include:
    Hazardous Materials Control  Research Institute
    U.S. Environmental Protection Agency
    U.S. Corps of Engineers
    U.S. Geological Survey
    Federal Emergency Management Agency
    American Society of Civil Engineers
    Association of State and Territorial Solid Waste Management Officials
    National Association of Local Governments on Hazardous Waste
    Centers for Disease Control
    Portland Cement Association
  The professionals on the Program Review Committee intensively reviewed hundreds of
abstracts to develop an informative and interesting agenda; the "best yet" according to those
who have read all the manuscripts. The Committee was composed of:
    Wayne Adaska, Portland Cement Association
    Hal Bernard, Hazardous Materials Control Research Institute
    John Farlow, U.S. Environmental Protection Agency
    Steven James, U.S. Environmental Protection Agency
    Brian Moran, U.S. Corps of Engineers
    Sue Perez, Federal Emergency Management Agency
    Barbara Simcoe, Association of State and Territorial Solid Waste Management Officials
    Jerry Steinberg, Hazardous Materials Control Research Institute/Water and Air Research
    Andres Talts, American Society of Civil Engineers/Defense Environmental Leadership
    Project
    Diane Van De Hei, National Association of Local Governments on Hazardous Waste
  A very special thanks to Dr.  Gary Bennett,  Professor of Biochemical Engineering, The
University of Toledo, Judy Bennett, Editorial Consultant, Toledo, and  Hal Bernard, Hazar-
dous Materials Control Research Institute, for editing this massive undertaking in the short
turnaround time. A more special thanks to the typesetters, graphics and proofreading team for
meeting impossible deadlines and to the HMCRI staff  for keeping it going in the right direc-
tion.

-------
                                               CONTENTS
                                                      ..6

                                                      .11

                                                      .16
            SITE DISCOVERY & ASSESSMENT

Feasibility Studies of Groundwater Pollution Source
Identification from Actual Field Monitoring Well Data ...
  Jack C. Hwang, Ph.D., Amin Ayubcha, Shih-
  Huang Chieh, Ph.D. & Stuart Richardson
Incorporation of Hydrogeologic Data into United
States Environmental Protection Agency/Environmental
Photographic Interpretation Center Investigations
of Hazardous Waste Sites	
  William L. Baer & Peter M. Stokely
Downtown Carcinogens—A Gaslight Legacy	
  Robert H. Salvesen, Ph.D.
An Underground Tank Spill Risk Assessment Program ...
  Daniel F. Predpall,  Warren Rogers, Ph.D. & Alan
  Lament, Ph.D.
                SCREENING TECHNIQUES

Detection of Groundwater Contamination by Shallow
Soil Gas Sampling in the Vadose Zone Theory
and Applications	20
  Eric G. Lappala & Glenn M. Thompson
Quality Control Attributes of Process Analytical Data	29
  Paul H. Friedman, Ph.D. & Duane Geuder
Characterization of Organic Wastes for Evaluation
of Remedial Action Alternatives	35
  Gregory A. Mooney & Russell W. Hartley
Chemical Composition of Drum Samples from
Hazardous Waste Sites	39
  William C. Blackman, Jr., Richard L. Garnas, Ph.D., John
  E. Preston, Ph.D. & Charlene M. Swibas
Survey of Mobile Laboratory  Capabilities and
Configurations	45
  J.L. Engels, H.B. Kerfoot,  D.F. Arnold,  R.H.
  Plumb, Ph.D. & S. Billets, Ph.D.
Construction of a Data Base From Hazardous Waste
Site Chemical Analyses	49
  Paul H. Friedman, Ph.D.,  William P. Eckel, Donald
  P. Trees & Bruce Clemens
Application of Mobile MS/MS to Hazardous Waste
Site Investigation	53
  David Ben-Hur, Ph.D., James S. Smith, Ph.D. &
  Michael J. Urban
                                                              Safety and Health Information for Use in
                                                              Responding to Hazardous Waste Emergencies
                                                               Jack Arthur
                                                              Methodology for Screening and Evaluation
                                                              of Remedial Technologies	
                                                                Virginia Hodge, Kathleen Wagner, Paul
                                                               Rogoshewski & Douglas Amman
                   AIR MONITORING

A Superfund Site Atmospheric Study: Application
to Remedial Response Decision-Making	
  T. laccarino, R. Stoner, R. Jubach &
  D. Smiley
The Effect of Wind Speed on the Emission Rates of
Volatile Chemicals from Open Hazardous
Waste Dump Sites	
  Jack Caravanos & Thomas T. Shen, Ph.D.
Air Monitoring at a Major Hazardous Waste
Cleanup Site: Objectives/Strategy/Results	
  John M. Bruck, Eugene W. Koesters &
  William R. Parker
Measurement of Volatile Organic Emissions
from Subsurface Contaminants	
  W. David Balfour, Bart M. Eklund & Shelly
  J. Williamson
On-Site Air Monitoring Classification by the Use
of a Two-Stage Collection Tube	
  Rodney D. Turpin, Kirit H.  Vora, J. Singh,
  A. W. Eissler & Daryl Strandbergh
                                                      .59
                                                      .62
.66
                                                                                                                   .68
                                                                                                                   .72
                                                                                                                   .77
                                                                                                                   .81
                                                                                    SAMPLING

                                                             Field Sampling for Monitoring, Migration and
                                                             Defining the Area! Extent of Chemical Contamination .
                                                               John M.  Thomas, Ph.D., J.R. Skalski, L.L.
                                                               Eberhardt, Ph.D. & M.A. Simmons
                                                             Subsampling of Hazardous Waste	
                                                               R. Swaroop, Ph.D. & O.S. Ghuman
                                                             Quality Assurance Audits of Field Sampling Activities
                                                               K. W.  Brown, D.S. Earth & B.J. Mason
                                                      .85


                                                      .90

                                                      .94

-------
                 LEACHATE CONTROL
                                                                          IN SITU & ON-SITE TREATMENT
Modeling Mobilization and Fate of Leachates Below
Uncontrolled Hazardous Waste Sites	
  Marcos Bonazountas, Ph.D. & Janet Wagner
Implementation of a Cooperative Agreement to
Investigate and Remedy Surface and Groundwater
Contamination at the Boulder/Marshall Landfill,
Boulder County, Colorado	
  Paul V. Rosasco & John Curry, Jr.
Investigations and Corrective Action: How It Was
Done at a Superfund Site in Connecticut	
  James Mack & Richard C. Dorrler
Evaluation of "Superfund" Sites for Control of
Leachate and Contaminant Migration	
  Robert S. McLeod
                BARRIER TECHNOLOGY

Laboratory Investigation of Physical Properties of
Flexible Membrane Liners (FML)	
  Brian D. Gish & William E. Witherow
Cement Barriers 	
  Wayne S. Adaska & Nicholas J. Cavalli
Barrier-Leachate Compatibility: Permeability of
Cement/Asphalt Emulsions and Contaminant
Resistant  Bentonite/Soil Mixtures to
Organic Solvents	
  David C. Anderson, Alicia Gill & Wayne Crawley
A Laboratory Technique for Assessing the In Situ
Constructability of a Bottom Barrier for Waste
Isolation	
  Thomas P. Brunsing, Ph.D. & Ray B. Henderson
                                                      .97
                                                     .103
                                                     .107
                                                     .114
                                                     .122

                                                     .126




                                                     .131



                                                     .135
                                                     .141
      CONTAMINATED GROUNDWATER CONTROL

Explosives Waste Disposal Sites: A DOD-Wide Problem
Case Study: Milan Army Ammunition Plant O-Line
Settling Ponds	
  Peter K. Wirth
Case History Organic Recovery and Contaminant
Migration Simulation	145
  Pressley L. Campbell, Ph.D.
Determining Contaminant Migration Pathways
in Fractured Bedrock	150
  Peter J. McGlew & J. Elliott Thomas, Jr.
NDT Location of Containers Buried in Saline
Contaminated Soils	158
  Robert M. Koerner, Ph.D. & Arthur E. Lord, Jr.,  Ph.D.
Contaminants in Groundwater: Assessment of
Containment and Restoration Options 	162
  A.C. Bumb, C.R. McKee,  Ph.D., J.M. Reverand,
  J.C. Halepaska, Ph.D., J.I. Drever, Ph.D. &
  S.C.  Way, Ph.D.
Control Technology Used in an Aquifer
Contamination Crisis Situation	170
  Jonas A. Dikinis, Kenneth J. Quinn & William D.  Byers
Control of Organic Air Emissions from
Groundwater Cleanup—A Case Study	176
  Abraham Thomas, Frederick Bopp, III, Ph.D.,
  John Noland, John Barone, Ph.D., Thomas
  Pierson & Michael Apgar
Inplace Closure of Previously Backfilled and Active
Surface Impoundments	
  Wayne Crawley, K. W. Brown, Ph.D. & David
  Anderson
Organic Sludge Stabilization: An Option that Works  .
  John M. Rademacher & Charles R. Hanson
In Situ Vitrification—A Potential Remedial
Action Technique for Hazardous Wastes	
  V.F. Fitzpatrick, J.L. Buelt,  K.H. Oma &
  C.L. Timmerman
                  ULTIMATE DISPOSAL

Treatment, Solidification and Ultimate Disposal
of Hazardous Waste Streams in Salt Formations ..
  Ray Funderburk
Disposal of Shock Sensitive/Explosive Chemicals
Utilizing Explosive Detonation with Open Burning
  Joseph S. Buttich & Samuel J. Gianti
Incineration of Explosives Contaminated Soils	
  John W. Noland A Wayne E. Sisk
An Overview of "Who Is Doing What" in
Laboratory- and Bench-Scale Hazardous Waste
Incineration Research	
  C.C. Lee, Ph.D. & George L. Huffman
.185
                                                                                                                   .189
                                                                                                                   .191
.195


.200

.203



.207
               FATE OF CONTAMINANTS

Methods of Determining Relative Contaminant
Mobilities and Migration Pathways Using
Physical-Chemical Data	
  Karl L. Ford & Paul Gurba
Endangerment Assessments for Superfund
Enforcement Actions	
  R. Charles Morgan, Robert Clemens, Barbara D.
  Davis, Thomas T. Evans, Jerald A.  Fagliano,
  Joseph A. LiVoIsi, Jr., Abraham L. Mittleman,
  J. Roy Murphy, Jean C. Parker A Kenneth
  Party miller
Migration and Degradation Patterns of Volatile
Organic Compounds	
  Patricia  V. Cline & Daniel R. Viste
                                                                          ENDANGERMENT ASSESSMENT

                                                             Site Assessment Under CERCLA: 'The Importance
                                                             of Distinguishing Hazard from Risk'	
                                                               Swiatoslav W. Kaczmar, Ph.D., Edwin C. T(fft,
                                                               Jr.. Ph.D. & Cornelius B.  Murphy, Jr., Ph.D.
                                                             The Importance of the Endangerment Assessment
                                                             In Superfund Feasibility Studies	
                                                               Anne Marie C. Desmarais d Paul J. Exner
                                                             Health Risk Assessments for Contaminated Soils	
                                                               Karl L. Ford & Paul Gurba
                                                             Public Health Significance of
                                                             Hazardous Waste Sites	
                                                               Robert L. Kay, Jr. & Chester L.  Tale. Jr.
                                                             Children's Exposure to Smelter-Associated
                                                             Lead, Montana and Idaho	
                                                               R. Schilling,  D.V.M., D. Ross, Sc.D., D. Sokal,
                                                               M.D., R. Ing, M.D., C. Brokopp, D.P.H. &
                                                               A.D. Maughan
                                                                                                                   .210
                                                                                                                   .213
                                                                                                                   .217
                                                      .221



                                                      .226

                                                      .230


                                                      .232


                                                      .239

-------
The Use of Serum Reference Materials and
Statistical Methods in the Classification of Human
Exposure to PCBs at Waste Sites	
  Vlrlyn W. Burse, John M. Karon, Ph.D.,	
  Douglas M. Fast, Ph.D. & John A. Liddle, Ph.D.

                   PERSONNEL SAFETY

Emergency Planning for Abandoned Waste Sites	
  Timothy G. Prothero, James Ferguson &
  William Martin
Guiding Principles  in the Design of Medical
Surveillance Programs for Hazardous Waste Workers	
  Edwin C. Hotstein, M.D.
Assessment of Human Health Effects from
Exposure to Environmental Toxicants	
  Karen K. Steinberg, Ph.D., Larry L. Needham, Ph.D.,
  John T. Bernert, Ph.D., James Myrick, Ph.D. &
  David D. Bayse,  Ph.D.
The Medical Surveillance of Hazardous Waste Workers	
  Frank L. Mitchell

                       SITE SAFETY

Multiple Chemical  Exposure Considerations
Associated with Health and Safety Assessments	
  Thomas C. Marshall, Ph.D.
Control of Fugitive Dust Emissions at
Hazardous Waste Cleanup Sites	
  Keith D. Rosbury & Stephen C. James
Safety Plans for Uncontrolled Hazardous
Waste Sites	
  /. Robert Steele
Heat Stress Monitoring at Uncontrolled Hazardous
Waste Sites	
  Rodney D. Turpin, William Keffer, Christopher Vias,
  Mary Helen Warden & John King

        RISK ASSESSMENT/DECISION ANALYSIS

Decision Making for Remedial Alternatives Using
the Provisions of CERCLA: PCB River Cleanup and
Industrial Site Cleanup/Closure	
  Lawrence M. Goldman & Roberta J. Fine
Effects of Uncertainties of Data Collection
on Risk Assessment	
  Gary L. McKown, Ph.D., Ronald Scholia &
  C. Joseph English
An Epidemiologic Study of Community Exposures
to 2,3,7,8-Tetrachlorodibenzo-p-Dioxin	
  Paul S. Stehr, Gary F. Stein & Karen Webb
The Application of Quantitative Risk Assessment
to Assist in Selecting Cost-Effective Remedial
Alternatives	
  Lawrence J. Partridge
Approaches to Computer Risk Analysis at
Uncontrolled Hazardous Waste Sites	
  Baxter Jones & Ken Kolsky
Coping with Uncertainty in Evaluating
Alternative Remedial Actions	
  William A. Tucker, Ph.D., Gregory J.
  Gensheimer & Robert F. Dickinson
Designing Risk Analyses to Avoid Pitfalls  in Cost
Recovery Actions: A Legal/Technical Solution	
  John C. Hall, J.D.
Practical Use of Risk Assessment in the
Selection of a Remedial Alternative	
  Katherine D. Walker & Christopher Hagger
                            COST OF CLEANUP
 .243
 .248
 .251
 .253
 .259
 .261
 .265
 .269
 .273
 .277
 .283
 .287
 .290
 .300
 .306
 .313
.3121
Cleanup Cost Allocation (CCA) Model	
  Richard B. Adams, Philip M. Zimmerman &
  Donald D. Rosebrook, Ph.D.
Remedial Cost Estimation System for
Superfund Sites	
  Kenneth T. Wise, Ph.D. & Paul Ammann
Cost Analysis for Remedial Actions Under Superfund
  Bruce Clemens, Edward Yang, Ph.D. & Brian
  J.  Burgher
A Revised Cost Management Approach for
Superfund Removals 	
  James R. Jowett & Robert J. Mason
Factors Influencing Cleanup Costs During a
Superfund Removal Action: A Case History	
  Lt. W.D. Eley & Lt. T.A. Baxter
                     POST CLOSURE

Statistical Considerations in Groundwater Monitoring
  Seong T. Hwang
Natural Resource Restoration/Reclamation
at Hazardous Waste Sites	,
  James R. Newman, Ph.D., Douglas S. McLeod &
  Jackson B. Sosebee
Assessment of Groundwater Contamination and
Remedial Action for a Hazardous Waste
Facility in a Coal Mine Region  	
  Mark J. Dowiak & Andrezej Nazar
Beneficial Reuses of Hazardous Waste Sites
in California	
  Julie K.  Anderson & Howard K. Hatayama
 .326



 .330

 .335



 .339


 .341
                                                               .346
.350
                                                               .356
                                                               .363
                PUBLIC PARTICIPATION

Public Information Needs in the Siting and
Cleanup of Hazardous Waste Sites	
  Nancy J. Jerrick & Nancy R. Tuor
Community Relations Activities for
Enforcement-Lead Superfund Sites	
  Daphne Gemmill & Bradley R. Brockbank
Managing Conflict in Controversial Siting
Issues: The Keystone Process	
  Nancy Ryburn Worst, Diane B. Sheridan &
  John Ehrman
Do Community Relations Matter?: The New
Jersey Perspective	
  Grace L. Singer
             ALTERNATIVE TECHNOLOGY

Electric Reactor for the Detoxification of
Hazardous Chemical Wastes	
  W.R. Schofield, Ph.D., J. Boyd, D. Derrington
  & D.S. Lewis
The Economics of Ground Freezing for
Management of Uncontrolled Hazardous
Waste Sites	
  John M. Sullivan, Jr., Daniel R. Lynch, Ph.D.
  & Iskandar K. Iskandar, Ph.D.
The Role of Adsorption and Biodegradation in
On-Site Leachate Renovation	
  Robert C. Ahlert, Ph.D., David S. Kosson, Erik
  A. Dienemann & Frederick D. Ruda
                                                               .368
                                                               .371
                                                               .374
                                                               .378
                                                               .382
                                                               .386
                                                               .393

-------
The Use of a Mlcrodisperslon of Air in Water for
In Situ Treatment of Hazardous Organics	
  Donald L. Michelsen, Ph.D., David A.
  Wallis, Ph.D. & Felix Sebba, Ph.D.
Thermal Treatment of Solvent Contaminated Soils	
  Diane Hazaga, Sue Fields & Gary P.  demons, Ph.D.
Floating Cover Systems for Waste Lagoons:
Potential Application at Old Inger Site, Louisiana	
  Mark L.  Evans, John P. Meade & Anthony N. Tafuri
Evaluation of Advanced Technologies for
Treatment of Contaminated Soils	
  Walter P. Lambert, Ph.D., Lawrence J.  Bove
  &  Wayne E. Sisk
Alternatives for Disposal of Unknown Gases,
Chemical Control Corporation Site,
Elizabeth, New Jersey	
  Dan Nickens, Jeff Gold & Henry Munoz
                     CASE HISTORIES

 Detoxification of Soils, Water and Burn Residues
 from a Major Agricultural Chemical
 Warehouse Fire	
   Mark D. Ryckman
 Superfund Removals in Remote Areas of the World:
 Pacific Island Immediate Removal Project	
   Michael H. Ridosh,  William E. Lewis, Christopher
   Vais &Lt.  Jack A. Kemerer
 Remedial  Investigation and Feasibility Study
 for the Pollution Abatement Services Site
 (Oswego,  New York) 	
   Daniel  W.  Rothman, John C.  Gorton, Jr. &
   James A. Sanford
 Case Studies  Involving the Treatment of
 Hazardous Substances under the Superfund
 Remedial  Action Program	
   William M. Kaschak &  James J. Spatarella
 Cleanup of Radium Processing Residues in a
 Highly Urbanized Area  	
   John M. Brink & Stephen F. Tarlton
 Superfund Planning Process for  the OMC
 Hazardous Waste Site, Waukegan, Illinois	
   Jack E. Braun & Stewart L. Davis
 Site Conditions and Corrective Action at the
 North Hollywood Dump	
   John D. Tewhey,  Ph.D., Andrew F. McClure
   & Terry K.  Cothron
 Radon Contamination in Monlclair and Glen Ridge,
 New Jersey: Investigation and Emergency Response ...
   John  K Czapor, Kenneth Gigliello &
   Jeanette Eng
.398
.404
.407
.412
.416
.420
.427
.435
.440
.445
.449
.452
.457
                   SITE REMEDIATION

Lessons Learned in the Conduct of Remedial
Action Activities	
  M. John Cullinane, Jr. & Richard A.  Shafer
Feasibility Study for Buried Arsenic Waste
at Perham, Minnesota	
  Catherine E. Jarvis, William E.  Thompson,  Jerry
  Rick & Deb McGovern
Enforcement Remedial Cleanup at the Petro
Processors Site; Baton Rouge, Louisiana 	
  Kevin G. Garrahan, Michael A. Kilpatrick &
  David E. Price
.465
.469
.478
Subsurface Geophysical Investigation and Site
Mitigation 	481
  Lee Taylor Lawrence
Guidelines for Decontaminating Buildings,
Structures and Equipment at Superfund Sites	486
  M.P. Esposito, J. McArdle, J.S. Greber &
  R. Clark
Contingency Planning for Remedial Actions at
Hazardous Waste Sites	489
  Robert Goltz, M. Shaheer Alvi & Salvatore
  Badalamenli
Selecting Superfund Remedial Actions 	493
  D. Brint Bixler A J. Bill Hanson
Guidance for the Conduct of Remedial Investigations/
Feasibility Studies under Superfund	498
  Richard L. Stanford, James Lounsbury, Douglas
  Ammon, S. Robert Cochran A Virginia Hodge
Cleanup of Radioactive Mill Tailings from
Properties In Monticello, Utah	504
  Floyd D. Nichols, John ,\f.  Brink A Philip C. Nyberg
Bidding Considerations for Superfund Cleanup
Contracts	509
  William F. Bonneau & Robert F.  Smart
The U.S. Army Installation Restoration Program	511
  Andrew W. Anderson A Lt. Colonel Paul E. Couture

                    STATE PROGRAMS

The Challenges of Siting Hazardous Waste
Management Facilities	517
  Joe Teller
Bid Protests, Change Orders and Claims in the
Superfund Remedial Program	521
  Linda Garczynski & Laurie M. Ziegenfus
S-Area: Negotiated Remedial Program for the
Niagara Frontier's Most Complex Site	525
  C. Kenna  Amos, Jr. & \furray E. Sharkey
Constructive Criticism on the Implementation of
the Superfund Program—A State Perspective	532
  Jim Frank
Implementation of RCRA Section 3012 at 160
Hazardous Waste Sites in Washington State	535
  Patricia M. O'Flaherty, Richard W dreiling &
  Barbara J. Morson
RCRA 30' 2 and Superfund Enforcement at
the State Level	544
  Charles R. Faulds, Daniel L. Scheppers &
  David Johnson
Hazardous Waste Policies and Management
Practices of New York City	546
  Carey Weiss, Stanley Siebenberg & Charles Smith

              INTERNATIONAL ACTIVITIES

The N ATO/CCMS Study of Contaminated Land	549
  M.A. Smith
Desecration  and Restoration of the Lower
Swansea Valley	553
  E.M. Bridges, Ph.D.
Investigation of Land at Thamesmead and
Assessment of Remedial Measures to Bring
Contaminated Sites into Beneficial Use	560
  George  W. Lowe
Remedial Action for Groundwater Protection
Case Studies within the Federal Republic of Germany	555
  Klaus Stief

-------
Review of the Development of Remedial Action
Techniques for Soil Contamination in the
Netherlands	569
  Dick Hoogendoorn
Extraction as a Method for Cleaning Contaminated
Soil: Possibilities, Problems and Research 	576
  W.H. Rulkens, Ph.D. & J. W. Assink
Measurement  of Low Permeability Coefficients
by Means of Electronic Instruments	584
  Wilhelm  Georg Coldewey, Ph.D.
The Upward Migration of Contaminants
Through Covering Systems	588
  R.M. Bell, Ph.D. & G.D.R. Parry, Ph.D.
Synopsis of 1983-1988 Outlook of Environmental
Concerns from Scientists Around the World	592
   William J. Lacy & Robert F. Holmes
Overview of Hazardous Waste Site Problems
in Wales	594
  Ronald A. Page
Treatment and Disposal of Hazardous Wastewater
from Foundry Cyanide Heat Treatment of
Delta Steel Company Limited	
  Edwin Ohonba & Sylvester Obaseki
Unification and Recycling of Groundwater
Contaminated with Petroleum Products and
Cyanides—The Karlsruhe (Federal Republic of
Germany) Drinking Water Treatment Plant	
  Rip G. Rice, Ph.D.
An Overview of Solid Waste Management in China..
  Gu  Youzhi & Zhu Yaohua
AUTHOR INDEX	606
SUBJECT INDEX	609

NATIONAL PRIORITIES LIST—
  Proposed Sites Update #2, October 1984	612

1984 Exhibitors	614
.598




.600

.604

-------
   FEASIBILITY STUDIES OF  GROUNDWATER POLLUTION
          SOURCE IDENTIFICATION FROM ACTUAL FIELD
                               MONITORING  WELL DATA

                                          JACK C. HWANG, Ph.D.
                                              AMIN AYUBCHA
                                       Department of Civil Engineering
                                               Drexel University
                                          Philadelphia, Pennsylvania
                                       SHIH-HUANG CHIEH, Ph.D.
                                          STUART RICHARDSON
                                       Ecology and Environment, Inc.
                                              Buffalo, New York
INTRODUCTION

  Locating a pollution source is perhaps the single most important
aspect of groundwater pollution, isolation and cleanup. Conven-
tional methods of locating an unknown source require an extensive
number of observation wells from which sufficient data can hope-
fully be obtained to plot concentration contours for identifying the
source location.
  Only recently, the use of numerical methods has been investi-
gated.  Computer modeling is helpful because use of  a limited
amount of well data allows one to back calculate to determine the
likely source.  This  process  is  called  "inverse  problem" or
"parameter  identification problem". An operative  computer
model requires only a few wells to effect a solution. The economic
implications of requiring only a few wells are obvious.
  Two different approaches were used for the numerical identifica-
tion: nonsequential and sequential optimization. For the former,
the optimization is applied to the partial differential equation of
groundwater solute transport.  For the latter, the partial differential
equation is spatially discretized and is rearranged into a standard
form of linear dynamic system. The optimization is then applied to
the dynamic equation.
  A typical nonsequential approach was described in a paper by
Gorelick et a/.1 in which the optimization methods of linear pro-
gramming and multiple regression were combined with numerical
simulation of groundwater solute transport to identify the loca-
tion and magnitude of groundwater pollutant sources.
  As a sequential approach, the  development ;of analytic formu-
lations  and methodology using sensitivity theorem was  described
in the author's  earlier paper.2 The computer codes based on the
algorithms so developed were applied to field situations. Two cases
of field application will be presented in this paper. For the sake of
clarity, the basic principles of the methodology will be outlined
first.

BACKGROUND
  The spatially  discretized partial differential equation of ground-
water solute transport using finite element method can be written
as:
    M {c} + (K-A+E)  {cj - |p} = 0
                                           (1)
in which
     {c} - (del
     I '   Utl
concentration vector representing the concentration

  the pollutant at each nodal point

     time derivative of the concentration vector
                                                  K, A, E and M = coefficient matrices, dispersion, convection,

                                                               first order decay and mass matrix respectively.

                                                 jpl • forcing vector
  If the boundary conditions prescribed are of Neumann type with
normal derivative equal to zero and there are isolated sources with-
in the domain confined by the boundary, it can be shown that the
forcing vector, (pj, has zero entries except at the "source" nodes
where the entry values are related to the magnitudes of the pollu-
tant at corresponding source nodes. Thus, the problem of locating
a pollution source from limited well data is equivalent to identify
the forcing vector [pj in Equation 1 for some given entry values in
the concentration vector [c] at some discrete instant in time.
  The sensitivity theorem for a linear dynamic system3 provides a
recurrent formula for updating  |pj in a manner so as to minimize
the error' (sum of the square of the differences between estimated
and measured concentrations at monitoring wells). The reader is
referred to the original paper2 for detailed derivations.
  The computer programs enable the user to identify the source
location of groundwater pollution using the existing limited data. If
the preliminary computer prediction is not decisive, it can suggest a
neighborhood for drilling locations of additional monitoring wells.
With additional new sets of concentration data as input to the com-
puter program, the prediction is expected to improve in the subse-
quent computer run. Thus, the neighborhood  of the suspected
source location is narrowed. By repeating the process, one may
eventually pinpoint the source location.
  Ecology and Environment,  Inc.,  provided  Drexel University
hydrogeologists information of the  site  including groundwater
table elevation, geological formation, well location and descrip-
tion, brief history of the site and laboratory report sheets of chem-
ical analysis of the groundwater samples from the monitoring wells.
The computer prediction made  by  the Drexel team was then
checked by Ecology and Environment, Inc. against the actual find-
ings from the field investigation—results which were not released
to the Drexel team  before running the computer programs. Two
cases of field applications will  be described in the following sec-
tions.

CURTISS WELL CASE
Description of the Site

  The Curtiss Well pollution Site  is located in  Southington,
Connecticut (Fig. 1). In the figure, the encircled area (1500 ft by
                                                                           SITE DISCOVERY & ASSESSMENT
                                                                                                        1

-------
                                                           Figure 1
                              Topography of the Studied Area and the Location of Wells in the Curtiss Well Site
4000 ft) is the domain of the numerical solutions. The area of in-
terest is mostly covered by a sandy, locally silty,  formation which
overlays the bed rocks.  The Quinnipac  River  runs north-south
halfway across the region and then turns southwest leaving the
area.
   Production well #4 (Fig. 1) has been in operation,  pumping at a
rate of 380 gal/min, since 1955. Due to the demand  for water, an
additional production well (well #6) on the other side of Quinnipac
River was developed. When the new production well began opera-
tion in 1976, its water was found to contain a high concentration of
organics. Similar results  were also  found in production well  #4
later. Both wells were capped in 1976.
   An initial investigation found that there was a surface ponding of
industrial organic waste during the 1950s and  1960s. This discovery
started the extensive site  investigations of the area.  A number of
monitoring wells were installed, and the results of chemical analy-
ses for the groundwater quality were compiled in 1980,  1981 and
1982.
   Among the dissolved organics in groundwater, TCE (trichloro-
ethylene) was chosen for the modeling of the groundwater solute
transport in the studied area. The measurements from the samples
at the wells with significant TCE concentrations were summarized
in Table 1. The locations of the wells are shown in Figure 1.

Solution Procedures and Results

  There are three separate programs in the computer package for
source  location identification. The first  program computes  the
nodal head values. The output of the first program  will  be incor-
porated in the input file to the second program which computes the
velocity components at each node. The values of nodal  velocities
are then used by the third program which predicts the distribution
of the strength of the groundwater pollution source.
   For the preparation of input files to these programs, one has to
set up a grid system  for the discretized site and determine several
model parameters. As shown in Figure 2, the Curtiss Well Site is
discretized into 96 quadrilateral elements with  119 nodes.
  A non-uniform but steady flow field was used. It was felt that
groundwater  table contours at pumping conditions (production
well #4 pumping at a constant rate of 380 gal/min) would be more
appropriate than that of natural conditions (no pumping) for this
source identification study. The depth of aquifer varies across the
study  site; the variation in depth is shown in Figure 3. It is assumed
that TCE is instantaneously mixed with water,  and the concentra-
                           Table 1
         TCE Concentration Measurements at Curtiss Well Site
Well
TCE
No. Node No
SW-5
WE-1
TW-8A.B
TW-7
TW-11
CW-1-78
CW-5-78
FWI6
CW-7-78
CW-6-78
CVI-5-78
CW-10-78
CW-8-78
PWI4
1
10
27
28
40
60
61
69
78
79
90
101
104
105
cone .
0.4
0.6
700K
270
25.5
1.8
0.1
1.0
120
7.7
1.7
7.0
7.0
4. 7
date
3/80
3/80
3/80
3/80
3/80
6/15/81
3/80
3/80
6/15/81
3/80
6/15/81
3/80
3/80
3/80
TCf
cone .
0.8
6.8
36. 5K
340
5
32
60
4.2

3.0
4.0
1.7
1.7
2.8
date
6/17/S!
6/17'»l
8/11/82
8/11/8:
9/15/8.
9/15/82
9/i5/f>:
7/7 '81

9/15/82
9/15/82
h/15/81
6/15/81
6/17/81
TCE
cone, date
210 8/11/6:
15 8 11/82
-OK 9/15/^




2.0 9/15/82



<0.1 9/15/82

7/7/81
                                                                 Note; All concentrations in M
       SITE ASSESSMENT & DISCOVERY

-------

                                                    110
                101
                              115
                                  116
       66
                     1000

            X-dls1ance
-------
                            Figure 4
               The Velocity Field in Curtiss Well Site
                           Figure 5
 The Predicted Source Locations and the Actual Finding from the Field
                        Investigations

  The Lowry municipal sanitary and industrial waste landfill is
located to  the  west of a military reservation. The landfill was
opened in 1964; in  1967, the City of Denver authorized dumping
liquid  hazardous wastes. It was  estimated  that  approximately
100,000,000 gal of the liquid wastes were dumped between 1967 and
1980.
  Significant concentrations of  hazardous  organics  began  to
appear in the groundwater samples in  several of the monitoring
wells. A  number of new monitoring wells were installed, and the

4       SITE ASSESSMENT & DISCOVERY
                                                                                             Figure 6
                                                                        Topography and the Location of WeMs in Lowry Landfill Site
                            Figure?
            The Velocity Distribution in Lowry Landfill Site

 results of chemical analyses for the groundwater quality were com-
 piled in 1981,1982 and 1983.
 Procedures and Results

   The procedures utilized in  this study are essentially the same as
 described in the  previous section: field discretization,  estimation
 of parameters  and  nodal velocity  computations. The parameters
 and  their values were listed in Table 2. and the velocity field was
 plotted in Figure 7. For  simplicity, other information will not be
 presented here.
   The computer predictions based  on TCE transport were plotted
 in Figure 8. The shaded area encircling nodes 59 and 60 is the pri-
 mary source location.  The secondary source area includes nodes
 58,65, 64, 77,70,66.72,67 and 54, and the other secondary source
 area is the surrounding of node 91. Since the groundwater is flow-
 ing from south to north and node 91 is located south to the primary
^TrpnM°d,eS,59/nd 6°'I1 is Unlikely that thc hi*h concentrations
of TCE detected near node 91 are  transported from nodes 59 and
60. It may be that the landfill covers a larger area. Indeed  it has
been reported that various industrial wastes were dumped at'differ-

-------
                                                                                                     Boundar? of predicted
                                                                                                     *r«* «ourc*
                                                                                                    Location of burial cclli
                                                                                                    found in the field
                           Figure 8
            The Distribution of Source Strength for TCE
ent locations within the large area landfill. To obtain better boun-
dary of the large area contaminant source, 1,1-Dichloroethane and
total organics were selected for the modeling in addition to TCE.
  The composite area source which was obtained by superimpos-
ing the predicted area sources for all  three species is shown in
Figure 9. Also, shown in the figure are the locations of burial cells
found in the field. As can be seen, the projected  sources fall rela-
tively within these areas.

CONCLUSIONS

  The general problem of locating a pollution source by having
data from a limited number of observation wells is quite challeng-
ing, yet worthwhile.  Major advantages are the  shorter time and
lower  costs involved compared  to the  conventional method  of
pollutant  contour  mapping.  It  has  been  demonstrated in  the
author's earlier paper2 that the proposed method can accurately
identify a pollution source location with a limited number of obser-
vation wells using hypothetical data.
  In this  study, the  test of feasibility of using these computer
models has  been carried one  step further by applying the corn-
                           Figure 9
     The Composite Area Contaminant Source and the Locations of
                          Burial CElls

puter codes to two sets of real field data: Curtiss Well Site and
Lowry Landfill Site. Based on the results, it can be concluded that
the computer program can be developed into a useful tool. It holds
the promise  of being  able to determine  where additional  wells
should be located for an on-going pollution scheme.

ACKNOWLEDGEMENT

  The project was  sponsored by the USEPA/Ecology and En-
vironment, Inc. under contract No. HQ-8307-04 and by a research
grant from the office of the Hazardous and Toxic Waste Man-
agement, Pennsylvania State University.

REFERENCES

1. Gorelick, S.M., Evans, B. and Remson, I., "Identifying Sources of
  Groundwater Pollution: An Optimization Approach,"  Water Re-
  sources Research, 19, 1983, 779-790.
2. Hwang, J.C. and  Koerner, R.M.,  "Groundwater  Pollution Source
  Identification From Limited Monitoring Well Data: Part 1—Theory
  and Feasibility", /. of Hazardous Materials, 8, 1983, 105-119.
3. Frank, P.M., Introduction to  System  Sensitivity Theory, Academic
  Press, New York, NY, 1978.
                                                                                    SITE ASSESSMENT & DISCOVERY

-------
      INCORPORATION OF HYDROGEOLOGIC  DATA  INTO
UNITED STATES ENVIRONMENTAL PROTECTION  AGENCY/
    ENVIRONMENTAL PHOTOGRAPHIC  INTERPRETATION
  CENTER INVESTIGATIONS  OF  HAZARDOUS WASTE SITES
                                           WILLIAM L. BAER
                                           PETER M. STOKELY
                                         The Bionetics Corporation
                                            Warrenton, Virginia
 INTRODUCTION

   Environmental  assessment using  historic  and  current  aerial
 photographs is the primary function of the USEPA's Environ-
 mental Photographic Interpretation Center (EPIC) in Warrenton,
 Virginia, a field station of the Environmental Monitoring Systems
 Laboratory, Las Vegas, Nevada. A major portion of EPIC's cur-
 rent workload is to  provide  photoanalytic support to  federal
 Superfund investigations. In response to the need for hydrogeo-
 logic data in these investigations,  EPIC has begun incorporating
 available hydrologic and geologic information into its reports.
 Two studies are presented which  illustrate EPIC's initial efforts
 in this area.
 NORTHWESTERN VIRGINIA SITE INVESTIGATION

   Concerns about the possible effects of an industrial plant in
 northwestern Virginia on the surrounding community and  en-
 vironment arose after a contaminated well and a highly acidic pond
 were discovered in the vicinity of the plant. An analysis of  his-
 torical aerial photographs was undertaken to document past activ-
 ities and conditions at the site. In addition, an effort was made
 to locate the plant site within its regional geologic context since
 hydrogeologic conditions in the  area were critical factors in  de-
 termining potential pollutant pathways.
   A search of  government and  commercial aerial photographic
 sources was undertaken to obtain the  best quality photography
 available of the site for the period 1945 to 1979. Black and white
 photography was obtained for 1945, 1950, 1958, 1964, 1970 and
 1976, and false color infrared photography was obtained for 1979.
 Sources of the black and white photography were the Agricultural
 Stabilization and Conservation Service (ASCS), the U.S. Geologi-
 cal Survey (USGS) and the Virginia Department of Transporta-
 tion. The false color infrared photography was obtained from
 EPIC's in-house film library.
   Each year of photography was analyzed for possible sources of
 pollution including tanks, ground stains, buildings, pipelines,  im-
 poundments, waste burial areas and on-site activities. The analysis
 was  performed  by stereoscopically viewing pairs of transparen-
 cies, backlit on  a standard light table. By observing the site three-
 dimensionally, and at various magnifications, the analyst could
 search  for objects, features and signatures associated with poten-
 tial pollution sources. A land use  and drainage survey of the study
 area was also performed.
  Enlargements were made from coverages which revealed signif-
icant changes in the study area. Findings were annotated on over-
lays to these enlargements, and full descriptions were provided in
an accompanying text.
  The type of information that was obtained from the analysis of
historical aerial photographs is shown in Figures  1 to 4.  In 1945
(Fig. 1), the plant is being constructed, renovated or converted and
a fill area (Fill 1) composed of earthen material and/or rubble is
visible. By 1950 (Fig. 2), the plant is fully operational. This can be
inferred, in part, by  the presence of a coal pile. The initial fill
area has been enlarged since 1945, and a new fill area (Fill 2) has
been started in a former field.
  By 1958 (Fig. 3), the second fill area has been greatly expanded
and small amounts of standing liquid can be seen on its surface
and borders. The first fill area does not appear to have received
additional fill material. In 1964 (Fig. 4), the plant does not appear
to be operating and no coal pile is present. The second  fill area
appears to have received additional material since 1958, and a large
pond has formed adjacent to it. This pond has recently been
determined to be highly acidic.
  Geologic information was obtained from the Virginia Division
of Mineral Resources and the U.S.  Soil Conservation  Service.
This information  included maps and accompanying descriptive
materials on the study area's soils, surficial geology and bedrock
geology. Overlays to photographic enlargements  were produced
using these sources.
  A portion of the photographic overlay that depicts soil  types in
the vicinity of the site is shown in Figure 5.' Deep and well-drained
loams and silt loams  underlie the site and have clay or silty clay
loam substrata (2B and 51C). Depth  to bedrock  and high water
table are greater than  1.5 and 1.8m, respectively. The fill areas and
pond identified in Figures 1 to 4 occur on these soils. The plant site
borders on another soil unit which is characterized by thin soils and
occasional bedrock outcrops (174B).
  A portion of the photographic overlay which depicts the  surficial
and bedrock geology  of the study area is shown in Figure 6.1-3% *
Surficial deposits  underlying portions of the plant site  include
permeable sand, clay and cobbles. Bedrock underlying the entire
site  is limestone and dolomite possibly interbedded with chert
masses, sandstone,  shale and/or conglomerate. In an unfractured
state, limestones and dolomites are relatively  impermeable to
water, but solution of bedding planes, joints and  faults may pro-
duce routes for rapid groundwater movements.
  By utilizing basic geological information in conjunction  with the
results of aerial photographic analysis, the field investigator has
       SITE ASSESSMENT & DISCOVERY

-------
     Figure 1
Industrial Site, 1945
     Figure 2
Industrial Site, 1950
     Figure 3
Industrial Site, 1958
     Figure 4
Industrial Site, 1964
                                                              SITE ASSESSMENT & DISCOVERY

-------
                          Figure 5
                          Soil Series
                                                                 quest for a historical aerial photographic inventory and analysis
                                                                 of the Upper Merion Township, located in southeastern Pennsyl-
                                                                 vania.
                                                                   The Upper Merion Township study area is a 3 km x 10 km rec-
                                                                 tangle in southeastern Pennsylvania incorporating Bridgeport, por-
                                                                 tions of Norristown, West Conshohocken,  Valley Forge National
                                                                 Historical Park and the suburban, industrial and commercial devel-
                                                                 opment in the area (Fig. 7).
                                                                   In the inventory, historical and current aerial photography were
                                                                 used to identify and describe sites that may contribute to ground-
                                                                 water contamination.  Sites were regarded as  potential  ground-
                                                                 water contamination sources if they had  been  used  for waste dis-
                                                                 posal or some other activity which may have had a negative impact
                                                                 on surface or  groundwater quality. These sites included quarries
                                                                 (mostly abandoned),  other old excavations, depressions, impound-
                                                                 ing basins, vacant lots, auto junkyards, industrial sites and land-
                                                                 fill sites. Available hydrogeologic data, consisting of aquifer yield
                                                                 and surficial geologic maps, were included in the analysis.
                                       METAMORPHCS
                          Figure 6
                 Surficial and Bedrock Geology
                                                                                     Figure 7
                                                                  Location Map, Valley Forge & Norristown, PA Quads
gained a significantly greater understanding of conditions at the
site prior to actually visiting the site. Thus, the investigator will
not only be able to determine the likelihood of groundwater  con-
tamination at the site but also will know what  to expect when a
site visit is conducted.

REGIONAL INVENTORY

  Investigations into waste disposal activity and the discovery of
traces of chemical contamination in a local reservoir led to the re-
                                                            The site information gathered from the analysis of the historical
                                                          aerial photography was compiled in textual and  map  form.  The
                                                          textual information is a description of the site as it appears on the
                                                          aerial photography. It includes information on site size,  the type of
                                                          site, location of solid and liquid waste  disposal areas, drainage
                                                          pattern and other environmentally significant features. This analy-
                                                          sis was done for each year of aerial photography so that the se-
                                                          quential development of each site could be understood. The loca-
                                                          tions  of  the  sites  as  seen on the  aerial  photography  were  then
                                                          transferred to a base map (Fig. 8).
8
SITE ASSESSMENT & DISCOVERY

-------
                          Figure 8
                        Site Locations

  Hydrogeologic information from previous groundwater studies
was obtained from the Commonwealth of Pennsylvania Depart-
ment  of Forests and  Waters and the Department of Environ-
mental Resources. This information included textual information
describing the properties of the geologic units, maps depicting med-
ium yield of wells drilled into  the  area aquifers and maps de-
picting local surficial  geology.  The maps showing the  medium
yield of drilled wells corresponded nicely to the surficial geology
maps of the area; a simplified map showing the combination of
this information was easily made (Fig. 9).
  All of the rock types found in the Upper Merion Township can
be expected to yield supplies of water and, therefore, are con-
sidered aquifers.  However, the only aquifers in the area  that can
be expected to yield large supplies of water are members of the
northeast-trending carbonate rocks (defined as sedimentary rocks
composed of limestone and dolomite) that underlie a large portion
of the study area. These carbonate  rocks not only yield large quan-
tities of water, but they are also  highly developed for public water
supply.5'6 The carbonate formations have been deeply weathered
and the  secondary openings greatly enlarged by solution. This
weathered zone has a higher porosity than the unweathered rock,
and where it does not contain large amounts of clay it may have a
relatively high permeability. The median depth of this weathered
zone is  12 to 23  m but can be as  deep as 41 m or as shallow as
8m.!
  Water in  this  zone occurs chiefly  under water table conditions
and is recharged directly from precipitation.5 Much of the ground-
water discharged to streams probably passes only through this
weathered zone, which is  usually an  important source of recharge
to the fractures in the underlying bedrock.5 Pollutants on the land
surface  or buried in this  zone may  be carried downward by in-
filtrating water to contaminate  the surface  water,  the weathered
zone and the underlying bedrock. The above is probably true of
many formations, but the enlarged secondary openings of the car-
bonate formations may speed the  movement of surface and sub-
surface water into the bedrock.
                                                                   The potential groundwater contamination site locations identi-
                                                                 fied from the aerial photography were overlaid on the aquifer in-
                                                                 formation maps (Fig. 10). When this was done, it became apparent
                                                                 which potential groundwater contamination sites fell above a par-
                                                                 ticular aquifer. This allowed the user to access the potential threat
                                                                 to each aquifer relative to its potential value as a water resource.
                                                                   Twenty of the sites are located in or just above these weathered
                                                                 carbonate aquifers. Within the study area, at least 38 quarries or
                                                                 excavations  have been opened into  these formations. Many of
                                                                 these have subsequently been filled. If these quarries were repos-
                                                                 itories for contaminated fill or hazardous waste, a direct conduit
                                                                 could exist for contaminants  to enter the groundwater supply of
                                                                 these highly utilized formations. In fact,  some wells in the carbon-
                                                                 ate rocks of the area have reportedly been abandoned because they
                                                                 are directly connected to polluted surface streams or other sources
                                                                 of pollution such as cesspools and waste disposal wells.6
                                                                   These  carbonate formations are especially susceptible to  con-
                                                                 tamination from waste disposal activity due to the deep weathered
                                                                 zone containing large secondary  openings  that  is overlain by a
                                                                 large number of potential contamination  sites.  In addition, the
                                                                 fact that these  aquifers are heavily developed for public supply
                                                                 makes waste disposal in them particularly dangerous.
 HELL YIELD LEGEND

fjj- High Yield (ZOO* gpm)
  - Medium Yield (110 gpm)


  - Low Yield (50 gpm)
    GEOLOGIC LEGEND

1 - Ledger Dolomite


\- Conestoga Limestone
d

: - Elbrook Formation (Limest
                   Figure 9
             Well Yield and Geology
                                                                                   SITE ASSESSMENT & DISCOVERY

-------
       HELL »IELO IEGEKD

      1- High Yield (200* gpai)
        - NedllM Yield (110 JP")


        - Lou Yield (90 gp-1
                              GEOLOGIC LEGEND

                            - Ledger Dolonlte


                            - Conestogi Llmttone


                            - El brook formation (Ll«e«tone)
                           Figure 10
                   Sites, Well Yield and Geology

  Other formations in the study area can be expected  to yield
usable quantities of water. However, the fewer potential  contam-
ination sites, the lack of large excavated openings and the absence
                                                                  of deeply weathered layers with enlarged secondary openings make
                                                                  these formations less sensitive to contamination.
                                                                    In conclusion, the combination of aerial photographic analysis
                                                                  with existing hydrogeologic data indicates that a large number of
                                                                  potential groundwater contamination sites are located in the area
                                                                  of carbonate aquifers and that these aquifers are particularly sus-
                                                                  ceptible to contamination.  Furthermore, contamination of these
                                                                  aquifers is particularly significant because they are fully utilized as
                                                                  public water supplies.
                                                                   CONCLUSIONS
                                                                     Since completion of the studies described in this report,  EPIC
                                                                   has undertaken other projects incorporating  hydrogeologic in-
                                                                   formation. These studies have  all involved the reproduction  of
                                                                   existing hydrogeologicai information on overlays to photographic
                                                                   enlargements or maps. It is hoped that this type of data presen-
                                                                   tation will augment the usefulness of both the hydrogeologicai in-
                                                                   formation and the aerial photographic analysis. The usefulness of
                                                                   incorporating  hydrogeologic   information   with   information
                                                                   gathered  from aerial photography lies in the  ability to locate and
                                                                   describe historical waste sites and show their relationship  to the
                                                                   area  hydrogeology. These efforts are not meant to replace tech-
                                                                   nical field investigations of hazardous waste sites, but are meant to
                                                                   aid in the initial evaluation of potential groundwater  pollution
                                                                   sources.
REFERENCES
1.  United States Soil Conservation Service. "Soil Interpretations Record,"
   (preliminary), District Office, Luray, VA, 1983.
2.  Allen, R.M.,  "Geology and  Mineral Resources of Page  County,"
   Virginia Division of Mineral Resources Bulletin 81,1967.
3.  Gathright, T.M.. "Geology of the Shenandoah  National Park, Vir-
   ginia," Virginia Division of Mineral Resources Bulletin 86,1976.
4.  Rader, E.K., Webb. H.W., "Geologic Factors Affecting Land Modifi-
   cation, Warren County, Virginia," Division of Mineral Resources Pub-
   lication 15, 1979.
5.  Biesecker, J.E.. Lescinsky. J.B., and Wood. C.R., "Water Resources
   of the Schuylkill River Basin," Commonwealth  of Pennsylvania De-
   partment of Forests and Waters, prepared cooperatively by the United
   States Department of the Interior Geological Survey, Harrisburg, PA,
   1968.
 6. Newport, T.G.,  "Groundwater Resources of Montgomery County,
   Pennsylvania," prepared by the United States Geological Survey Water
   Resources Division, in Cooperation with the Pennsylvania  Geological
   Survey, 1971.
 10
SITE ASSESSMENT & DISCOVERY

-------
     DOWNTOWN CARCINOGENS—A  GASLIGHT  LEGACY

                                       ROBERT H. SALVESEN, Ph.D.
                                         Robert  H.  Salvesen Associates
                                              Red Bank, New  Jersey
INTRODUCTION

  Gas lights bring back memories of old-world charm and bygone
days. The source of gas to power lights was so-called town gas
manufacturered by a coal carbonization process during which by-
products were formed: coke, tar and ammonical liquor. The by-
product tar is the source of current concerns.
  In many areas where gas or coal tar were  manufactured, the tar
was stored in tanks or dumped into pits. Unfortunately, extensive
soil and  water contamination has occurred  through spillage and
leakage.  Much of the coal tar  and oils were used as a  source of
chemicals, to oil roads, as extenders for asphalt, as creosote to im-
pregnate wood and  as  fuel. Thus, these tars  can  be  found in
widespread areas.
  However, the major problem appears to be at the former sites of
the coal gas or tar plants where high concentrations of residues may
still be present as trapped pools of oil and tar or spread out over
considerable areas, contaminating soils as well as underground and
surface  waters.   The  major  contaminants  are polyaromatic
hydrocarbons (PAHs), many of them  carcinogenic, phenolic  com-
pounds,  heavy  metals and cyanides. These materials constitute
serious health problems.
  More than 1100 sites have been identified in the United States,'
and only a few have been examined in detail. Even fewer have been
treated to reduce or eliminate the potential hazards. The vast ma-
jority of the former sites have not been properly located, cleaned
up  or tested to determine the extent and nature of contamination;
nor are there plans to do so. In this paper the author describes the
history, gives examples of the specific problems and makes recom-
mendations  for generalized actions which need to  be taken at
former coal gas and tar sites. The seriousness of this problem could
dwarf many of  the currently  designated  Superfund  sites in
magnitude, since most sites are in downtown areas; they have not
been properly located and contain toxic materials. These sites con-
stitute serious health  hazards to an unsuspecting population.

BACKGROUND
  Before the availability of natural gas, the major source of  com-
bustible gas was coal. Following the lead of Great Britain, city coal
gasification plants were built in the United  States:  Baltimore in
1816; Boston in 1822; and New  York in 1825.' By 1920, there were
1,114 gas plants in the country.
  Gas was manufactured by heating coal white-hot  and pouring
water or oil over it to produce gas, tar and other products. The gas
was piped to the community to be used for illumination, heat and
industrial purposes.
  Some plants maximized production of coke and tar. Thus, there
were three types  of plants based on slightly different variations,
i.e., coke, gas and tar plants. All plants produced tar as a major
product or by-product, and this chemical is the prime concern of
this paper.
  Prior to 1887, when tar distillation was established as a separate
industry in Philadelphia, it can be assumed the tar produced was
largely disposed of at or near the plant site as a matter of con-
venience. The tar generated in the various plants was placed in
tanks or on the ground in pits or other locations. Because of these
disposal practices and the toxicity of coal tars, many of these sites
present serious environmental hazards.2'9 One  site in Burlington,
Vermont, appeared on the USEPA's list of 115 priority Superfund
sites;  subsequently, several others have been included on the ex-
panded list of 418 priority  Superfund sites. However, the vast ma-
jority of these sites have not been given adequate attention. Those
that have been dealt with have generally been discovered due to the
appearance of a problem. To illustrate this,  brief  histories of
several well-known locations follow.
Stroudsburg, Pennsylvania
  As a consequence of Hurricane Diane in 1955, Brodhead Creek
experienced its greatest  flood.  Subsequently, the Army Corps of
Engineers modified the stream channel between 1958 and 1960. By
1980, this  new stream bed was undercut about 6 ft; to strengthen
the levees  that had been built, some additional digging was per-
formed. In the course of this work, coal tar was identified in open
trenches along the shore of this creek. An investigation determined
that a coal gasification site had operated in this area for nearly 100
years prior to 1939.
  Approximately 10,000 gal of oil  were found in underground
pools  at the site, and extensive soil contamination was found be-
tween the plant and the creek, a distance of about 375  ft. Analyses
of the residual oil showed 15 Polynuclear Aromatic Hydrocarbons
(PAH), some of them known carcinogens i.e., Benzo (a) pyrene, at
concentrations ranging from 0.10% to 3.6%. Low concentrations
of phenolic materials were also found, but this is not unexpected
due to the extensive leaching of these water-soluble components
since 1939. High concentrations of metals such as Al (218 mg/1), Fe
(460  mg/1) and Mn (25.5 mg/1) plus  cyanides (0.30 mg/1) were
detected in shallow groundwater.
  This site received Superfund monies. It has been cleaned up, a
700 ft bentonite-cement slurry cut-off wall has been constructed to
prevent  further movement of pollutants  into  the  stream  and
monitoring wells are in place.

Ames, Iowa1

  Since 1927, aromatic hydrocarbons in jig/1 concentrations have
caused taste and odor problems in  the aquifer supplying Ames,
Iowa. This contamination has forced the abandonment of five city
wells close to the source and partial use of other wells. The source of
contamination, a gas plant waste pit abandoned in the late 1920s,
was identified in 1961 and removed to the sanitary landfill. How-
ever, problems with water  quality continued, and in 1975 an over-
                                                                              SITE ASSESSMENT & DISCOVERY
                                                                                                                       11

-------
flow channel (which flowed from the waste pit) was located as the
primary source of contamination.
  Tests showed the presence of 15 major classes of organic com-
pounds typical of coal tar residues in municipal wells. Aromatic
hydrocarbon  contamination was found in the sand  and gravel
deposits which form the "buried channel aquifer" from which
Ames obtains its water. Movement of coal tar  residues from the
original pit and surrounding areas was traced and proved by exten-
sive drilling, excavation and test work.
  After thorough study of the problems and potential solutions, a
pumping trough barrier was built to pump out the  contaminated
areas over the next 3-5  years. By that time, it is hoped the soil will
have leached  sufficiently to reduce the level of contaminants to ac-
ceptable  levels.  In addition, limited  pumping of wells has  been
initiated  and new wells have  been  drilled away from the  con-
taminated area.

Pittsburgh,  New York'
  From 1896 to  1960, a coal gasification plant operating along the
Saranac River in Plattsburgh deposited coal tar in unlined ponds at
this 11 acre site. Since 1960, periodic release of coal tar into the
river has been observed. Over the years, the coal tar ponds  were
filled  with  various  materials  including  ash,  cinders   and
miscellaneous soils.
  Soil to a depth of 13 ft (to fill or bedrock) beneath the original
ponds contains coal tar components. The heavier-than-water com-
pounds in coal tar formed a separate phase and moved along the
impermeable till  through sand and gravel to the river. Discharges of
coal tar into the river  were  sporadic and occurred  mainly in the
summer. The soil contained an average of 1.5% coal tar with con-
centrations ranging up to 9.6% at some locations. Detailed analyses
of the hydrocarbons were not conducted but may be presumed to
contain carcinogenic PAHs. High concentrations  of heavy metals
and phenols were also  found.
  Models were developed to study movement of coal tar with time
to aid in remedial actions. To prevent further contamination of the
river, soil along the  river bank  was excavated,  a  cofferdam  built
and uncontaminated soil filled in behind the structure. To prevent
further flow  into the river from inland areas, a cement bentonite
wall was constructed through the clean fill adjacent to the  river
bank. A 735  ft soil bentonite wall was built around  the main coal
tar pond site  and spoils area, covered  with a 36 mil Hypalon liner,
sand, topsoil and then  seeded.  Monitoring wells have been placed
at critical sites and  land use restrictions mandated by the New York
State Department of Environmental Conservation.

St.  Louis Park, Minnesota5'6'7

  Even prior to the closing in 1972 of a coal gasification plant that
operated for about 50 years, state and local agencies had been con-
cerned about  water quality in St. Louis Park, Minnesota. Based on
soil and water tests over an area of several square  miles, and at
depths to 700 ft, seven municipal water wells have been closed and
several others threatened. Twelve PAHs identified as carcinogens
by the USEPA were found in water and soil samples.  Concentra-
tions in the aquifers tested ranged from 30-200/ig/1 for known  car-
cinogens; other PAHs were found at levels of 200-3,000 /»g/l. The
USEPA's  Ambient Water Quality  Criteria  for Polynuclear
Aromatic Hydrocarbons, published in October 1980, recommends
zero as the allowable level.
  For the purposes of  the St.  Louis  Park studies, the following
criteria were established: (1) for individual  PAH identified as car-
cinogens, the assumed criteria was 2.8 /tg/1 (or the detection limit, if
higher) and (2) for other individual PAH, the assumed criteria was
28 /*g/l (or the detection limit, if higher). These criteria were ap-
plied to potable water and ambient groundwater.
  The extensive contamination of soil (much of it peat with a high
adsorbent capacity for organics) and aquifers could potentially
cause high concentrations of PAH for many years to come. Correc-
tive measures proposed, such as treatment with activated carbon,
have not been totally adequate. Since the USEPA and other agen-
                                                        cies have not determined acceptable levels for PAH, if there are any
                                                        in potable water,  the  problems at  St. Louis Park  remain unre-
                                                        solved. Control of pumping rates and drilling of new wells in un-
                                                        contaminated areas are among current considerations.

                                                        Ft. Lewis, Washington*

                                                           In December 1979, the U.S.  Department of Energy reported a
                                                        spill of about 2,300 gal of SRC  liquid during transfer of the liquid
                                                        from a storage tank to sample drums.' To prevent possible con-
                                                        tamination of ground,  surface and drinking water, a large volume
                                                        of soil was removed from the spill area. Soil to a maximum depth
                                                        of 20 ft was removed and replaced  with clean material. The land
                                                        surface was sealed and  wells  were installed to  pump  off  con-
                                                        taminated water  and  for  long-term  monitoring of  the  area.
                                                        Analyses  of the oil spilled showed it to  be a fuel oil blend, 2.1:1,
                                                        Middle Distillate: Heavy Distillate. Detailed analyses are provided
                                                        in the  report' and  may be summarized as follows:
                                                                                         Found by
                                                                                         Extraction
                                                        Aromatic & Aliphatic
                                                           Compounds
                                                        Phenolic Compounds
                                                        Basic Compounds
                                                             TOTAL
 82
 II
 6.9
99.9
              Found by
               GC-MS
 52
 5.4
 2.3
59.7
                                                           Approximately a dozen Priority Pollutants were found in  the
                                                         SRC fluid by GC-MS analysis. They  were: naphthalene, acenaph-
                                                         thene, fluorene, fluoranthene, phenanthrene/anthracene, pyrene,
                                                         chrysene/benzo (a), anthracene, benzol (b  + k), benzo (a  +  c)-
                                                         pyrene and phenol.
                                                           These materials were carried off in the removed soil and placed in
                                                         a secure landfill area. The remedial actions taken in this SRC fluid
                                                         spill have apparently localized groundwater contamination and
                                                         prevented intrusion into drinking  water supplies.

                                                         CURRENT CRISES-
                                                         NEW JERSEY AS AN EXAMPLE

                                                           Pollutants  from  over  1,100  former  coal  processing sites
                                                         throughout the country still present potential serious hazards to our
                                                         health and  environment. New  Jersey  (one  of the  most densely
                                                         populated states, with expanding industry and residential areas) has
                                                         done more  to  identify and  resolve this problem than any other
                                                         state. And  yet, with  nearly  60 sites identified (Table 1) by  the
                                                         Department of Environmental Protection," only one has been fully
                                                         tested. On about half of the sites,  coal tar residues have been iden-
                                                         tified; the remainder have had little or  no testing. The exact loca-
                                                         tion of several sites is not known. The most completely tested site is
                                                         at Belmar Township in Monmouth County.
                                                         BELMAR. N.J. GASIFICATION PLANT SITE

                                                           In August 1982, an oily discharge was discovered at a marina in
                                                         the nearby Shark River." Investigations traced the source of oily
                                                         discharge to the catch basin system for the  Borough recreational
                                                         park and garage.
                                                           It was determined that this site had been used for about 50 years
                                                         as a coal gasification plant. In 1952, it had been sold by Jersey Cen-
                                                         tral Power &  Light Co.  (JCPL) to New Jersey Natural Gas Co.
                                                         (NJNG). In 1971, all equipment, tanks and buildings  at the  site
                                                         were dismantled and NJNG deeded the site to the Borough in 1976.
                                                         Subsequently,  the Borough built a recreational park on this land.
                                                         During construction, the Borough hauled in approximately 1 to 5 ft
                                                         of fill material to cover the rubble located on the site and installed
                                                         three catch  basins along the northern  site boundary to improve
                                                         drainage.
                                                           In August 1983,  samples from the  soil borings and a nearby
                                                         stream were taken at the  Belmar site and analyzed (Table 2). Only
                                                         the base neutrals were separated for analysis. These results showed
12
SITE ASSESSMENT & DISCOVERY

-------
                                                                      Table 1
                                           Identified Coal Gasification Plants in the State of New Jersey
A. South Jersey Gas Company*
 1. Atlantic City—Kirkman Boulevard
 2. Atlantic City—Florida, Georgia and Sunset Avenues
 3. Pleasantville—Franklin Avenue
 4. Egg Harbor—Atlantic and Buffalo Avenues
 5. Hammonton—Twelfth Avenue
 6. Bridgeton—Vine and Water Streets
 7. Millville—North Second Street
 8. Glassboro—Union and Grove Streets
 9. Paulsboro—Jefferson Street, east of Billingsport Road
10. Swedesboro—Auburn Road and Bridgeport Road
11. Penns Grove—Pitman Street and the railroad tracks
12. Salem—Fifth and Howell Streets
•Present owners of former coal gasification plant
B. New Jersey Natural Gas Company/Jersey Central Power & Light
 1. Dover in  Morris County—Carrol Street
 2. Belmar—16th and railroad
 3. Cape May City—Lafayette and St. John Streets
 4.  Ocean City— llth and West (Atlantic  City Electric)
 5.  Long Branch—Long Branch Avenue and Brook Street
 6.  Lakewood—Clover Street and Laurel  Avenue
 7. Toms River—Water Street
 8.  Wildwood—West Garfield and Lincoln Avenue
 9.  Asbury Park—Prospect and Sewell
10.  Atlantic Highlands—W. Lincoln, Garfield and West Avenues
11.  Boonton
C. Elizabethtown Gas
 1.  Elizabeth—Erie Street between Third  and Florida
 2.  Elizabeth—South Street and Center Street
 3.  Perth Amboy—Margaret Street
 4.  Rahway—intersection of Central, Hamilton, Irving
 5.  Flemington—E.  Main Street
 6.  Newton—Driller Avenue
 7.  Lambertville—S. Main at Ferry Streett
                                                                               8. Washington Boro (Warren)—S. Lincoln at railroad trackst
                                                                               9. Newton—E. Clinton Avenuef
                                                                               10. Phillipsburg—Railroad tracks at Reese Courtf
                                                                               tSite located in the service territory of Elizabethtown Gas, but never owned or operated by the
                                                                               company.
                                                                               D.  Public Service Electric  & Gas
                                                                               1. Hobart Avenue Gas Works—Hobart and Oak Street, Bayonne
                                                                               2. Camden Gas Plant—Front and Spruce Streets, Camden
                                                                               3. Camden Coke Plant—Front and Delaware River, Camden
                                                                               4. Gloucester Gas Works—Jersey Avenue  and Sixth Street, Gloucester
                                                                               5. Hackensack Gas Works—Hudson and Water Streets, Hackensack
                                                                               6. Harrison Gas Plant—4th Street and Passaic Avenue, Harrison
                                                                               7. Hoboken Gas Works—13th and Clinton Streets, Hoboken
                                                                               8. Halladay Street Works—Halladay Street, Jersey City
                                                                               9. Old Provost Street Works—6th and Provost Streets, Jersey City
                                                                               10. West End Gas Plant—St. Pauls and Duffield Avenues, Jersey City
                                                                               11. Mount Holly Works—W. Washington Street, Mount Holly
                                                                               12. Front Street Works—McCarter Hwy. and Passaic River, Newark
                                                                               13. New Brunswick Works—Catherine and  Somerset Streets, New Brunswick
                                                                               14. Paterson Gas Plant—E. 5th and Leon Streets, Paterson
                                                                               15. Plainfield Gas Works—E. 4th and Washington Streets, Plainfield
                                                                               16. Central Gas Plant—Raritan River and Silver Lake Avenue, Edison Twp.
                                                                               17. Ridgewood Gas Works—Ackerman Avenue and Bellair Road, Ridgewood
                                                                               18. Riverton Works—Main Street, Riverton
                                                                               19. South Amboy Gas Works—George and Feltus Streets, South Amboy
                                                                               20. Trenton Gas Plant—Brunswick Avenue, Trenton
                                                                               21. Trenton Gas—365 South Warren Street, Trenton
                                                                               22. Woodbury Works—WFSSRR and Woodbury Creek,  Woodbury
                                                                               E.  Other Sites
                                                                               1. Kearney—Koopers Coke
                                                                               2. Tuckahoe
                                                                               3. West Paterson—Memorial Drivet
                                                                               4. Hawthorne—Route 208 NorthJ
                                                                               5. Hawthorne—Wagaraw RoadJ
                                                                               (Disposal sites
                                                                    Table 2
                                               Data from Belmar Site of Coal Gasification Plant
Date of Sample
Sample Number
Compu/Chem Number
Location





Base Neutrals
Units
ncenaphthene
Acenaphthylene
Anthracene
Benzo( a) Anthracene
Benzo(a)Pyrene
3,4 Benzofluoranthene
Benzo(k) f luoranthene
Bis(2-Ethylhexyl)phthalate
Chrysene
Bluoranthene
Fluorene
Phenanthrene
Pyrene
Naphthalene
Benzo(CHI)Perylene
Indeno(l ,2 ,3-cd) Pyrene
Anthracene/ Phenanthrene
Benzo(a)Anthracene/chrysene
DioChylphthalate
D irae thy Iphtha late
Di-n-butylphthalate
8/3/83
C83-62
10134
Stream Sedi-
ment by Park,
Boring #4,
About 300 ft
West of Aban-
doned Tank

ug/kg
4000
6800
6400
11000
9200
13000
13000
6800
11000
16000
6000
24000
23000
BDL (1)
BDL
BDL
-
-
-
-
-
8/3/83
C83-64
10136
On Site Soil
Boring #1, at
Base of Aban-
doned Tank



ug/kg
BDL (2)
300000
140000
110000
140000
85000
85000
BDL
100000
170000
260000
580000
240000
1300000
BDL
BDL
-
-
-

-
8/3/83
C83-63
10135
On Site Soil
Boring #3,
About 100 ft
NW of Aban-
doned Tank


ug/kg
BDL (3)
1500
BDL
1100
1100
1700
1700
BDL
1000
1400
BDL
520
1700
300
640
760
-
•
-
-
-
8/3/83
C83-65
10137
On Site Soil
Boring #2,
About 100 ft
East of Aban-
doned Tank


ug/kg
BDL (1)
5600
BDL
BDL
BDL
BDL )
BDL )
BDL
BDL
BDL
4400
9200
5200
29000
BDL
BDL
-
-
-
-
-
8/3/83





On Site Soil
Borings


ppm
BDL-0.37
11 0.04


0.01-8.9
0.01-4.6

BDL-0.07

0.02-3.7
BDL-0.09

0.02-9.7
0.02
BDL-4.6
BDL-3.2
BDL-0.73
0.0/-5.7
BDL-0.02
BDL-0.17
BDL-0.39
BDL «  Below Detection Limit
(1) =  4000 ug/kg
(2) -  50,000  ug/kg
(3) -  200  ug/kg	
                                                                                               SITE ASSESSMENT & DISCOVERY
                                                                                                                                                   13

-------
                                                           Table 3
                               Data from Belmar, N.J. Clothing Factory Adjacent to Coal Gulflcatlon Site
Dale of Sample
Sample Number
Sample Location




Base NeuCrali , ppm
Acenaphthene
Acenaphthylene
Anthracene/phenanthrene
Benzo (b ,k) f luorinthrene
Benzo (a) pyrene
Butyl benzylphthalate
Benzo (g,h,l) perylene
Bis (2-ethylheXyl) phchalace
Chrysene
Dibenzo (a ,h) anthracene
Di-n-butylphthalaCe
Diethylphchalate
DimeChylphthalate
2,4 Dlnitrotoluene
2,6 Dinltrocoluene
Fluoranthane
Fluorene
Indeno-(l ,2 ,3-c ,d) pyrene
Naphthalene
Pyrene
3/1/83
20056
Soil Sample
@ 5'3", 20 fl.
Eaal of Oil
Storage Tank
on Site

3.3
3.3
23.9
3.3
4.3
-
3.3
-
4.9
3.8
16.3
6.0
8.7
6.5
28.2
2.7
3.8
6.0
2.2
6.0
3/1/83
20058
Soil Sample
@ 4'1", 60 ft.
Weil of Oil
Storage Tank
on Site

913
870
1333
1130
1406
-
-
-
1648
-
-
-
-
-
-
493
449
-
-
681
3/1/83
20059
Inlet to Samp
Pump Pit In
Basemen!



19
12
34
5
6
-
4.4
66
8
-
8
-
-
-
-
7
IS
4.8
29
20
3/1/83
20057
Soil Sample
<& 4'10", 30 fl.
Sooth of Oil
Storage Tank
on Site



14.9
8
6.9
109
8
149
10.3
-
236
-
40


i.6
-
<..6
-
2.9
     Volatile Organic*
     Methylene Chloride
     Benzene
     Toluene
     M-Xylene
                                 150
                                                                                                      145
                                                            500
                                                            505
                       140
                         2
                        13
 the presence of a number of hydrocarbons typical  of coal tar
 residue.  Significant  concentrations of known  carcinogens and
 priority pollutants were found. Considerable variations were noted
 depending upon sampling location, with the highest concentrations
 generally found at the base of the abandoned tank. Very high con-
 centrations of carcinogenic and priority pollutants were also found
 in the stream bed sediment. This would seem to indicate  a buildup
 had occurred over a period of years, caused by a flow from the site.
  Based on the data obtained to date, there is significant cause for
concern due to the presence of known carcinogens. The site has
been closed down and fenced off to prevent access. The full extent
of soil and water contamination is not known, but more testing and
monitoring are currently underway by  JCPL.
  Adjacent to the Belmar site is a clothing factory which has had a
flow of oil into a basement sump for a number of years. Soils and
liquid from the sump have been tested  (Table 3). Again, only base
neutral extracts were examined. These data also show a wide range
of organic compounds, among which are carcinogens and priority
pollutants. Due to the location of an on-site fuel storage tank, these
data do not  show adequate proof that the coal gasification site is
the sole source contributing to this problem.  However, it is a good
possibility, and further analyses of collected  oils and  testing at ad-
ditional sites are  needed.

MAGNITUDE OF THE PROBLEMS
IN THE UNITED STATES

  Throughout the country, relatively little work has been done on
former coal gasification sites. As noted, there were 1,114 operating
coal gasification sites in 1920. Based on the New Jersey experience,
this number could be low by as much as 50% of the final total of
plants, since the industry  peaked in the 1940-50 era.  (1920 data
showed 36 sites  in New Jersey; recent studies  located almost 60
sites.)
  Inquiries have  been made to all USEPA Regions, state en-
vironmental  agencies and a number  of gas companies. Responses
from 29 states, the USEPA and two gas companies have identified
40 known sites outside the 60 found in New Jersey. Thus, over 90%
                                                         of the  former sites  are  unknown to the USEPA  and state en-
                                                         vironmental agencies.
                                                           On most of the sites investigated to date, serious health and en-
                                                         vironmental hazards have been found, and it may be anticipated
                                                         that the large number of unknown sites still present similar hazards
                                                         to an unsuspecting public. The major problems are to locate these
                                                         sites, assess the  hazards and take  remedial actions.  These are
                                                         discussed below.
                                                         No. of States
                                                         Contacted
                                                         50
              Nature of
              Responses
No. of Sites
Located
100*
0
No. of Sites Re-
ported in 1920 by
USGS in These
States
437*
437 1
              10 positive
              19 negative
'Includes 60 in New Jersey
tIdentical numbers arc coincidental
                                                         Discovery of Former Coal Gasification Sites

                                                           Of the six examples cited earlier, most were discovered by acci-
                                                         dent. Few governmental agencies have made attempts to look into
                                                         this problem for two major reasons: (1) these sites are not readily
                                                         apparent and records showing their existence  are difficult to ob-
                                                         tain; and (2) most states do not classify coal gasification residues as
                                                         hazardous wastes  and thus have  not established this as a high
                                                         priority concern.
                                                           Inquiries  made to 50 state environmental agencies have not pro-
                                                         duced the desired results:
                                                           Thus, even in the ten states which have initiated studies, all but
                                                         New Jersey have found less than 10% of those reported by USCS.
                                                         The reason  for this is probably that the data are not easy to find. A
                                                         review of the USEPA's ERRIS lists can provide leads, but the most
                                                         complete data resides with the gas companies. Even they have to do
                                                         considerable searching to find the sites. According  to the USGS
                                                         data, there were 960 companies producing gas from 1,114 plants in
                                                         1920;  almost  every plant  was separately owned. The current gas
14
SITE ASSESSMENT & DISCOVERY

-------
companies often inherited these operations. Over the years, most
plants were dismantled and the properties (often in what are now
prime  downtown areas) were used for other purposes or sold. At-
tempts to obtain data from gas companies have met with mixed
results,  generally unsuccessful. However, in order to fully resolve
this problem, cooperation from the gas companies is essential.
 Potential Health and Environmental Hazards

   Evidence  presented herein and  elsewhere has shown that coal
 gasification  residues do present serious health and environmental
 hazards. These problems need to be resolved. Extensive work has
 been done in England12'13 to identify and solve similar problems of
 soil and  water pollution.  To adequately identify and overcome
 these problems, consideration needs to be given to the following
 hazards:
 •Intrusion of carcinogens, PAH, phenols, heavy metals and cya-
  nides  into aquifers and water supplies must be controlled to ac-
  ceptable levels to protect our water supply
 •Skin contact with soil containing coal gasification residue may
  result in irritations and possible carcinomas
 •Ingestion and inhalation
 •Uptake of contaminants in food plants with heavy metals being a
  major concern
 •High concentrations of aromatic and unsaturated organics as well
  as heavy metals can inhibit or prevent plant growth
 •Chemicals  may attack  building materials and services. Hydro-
  carbons can migrate through plastic pipe and cable coverings as
  well as joint sealing compounds, possibly causing deterioration of
  these  materials. Acidic phenols and heavy metals can accelerate
  corrosion  of metals and concrete. Liquid residues are known to
  penetrate cement and cinderblock walls.
 •Both surface and underground fires and explosions are possible in
  areas  of pooled oil and tar residues. Careless ignition sources or
  underground short-circuits are several possible means of starting
  fires.
   It has been reported14  that epidemiological studies in Kentucky
 have shown high incidence of skin cancer among people in contact
 with  soil from a former coal gasification  site.  In Pennsylvania,
 buildings and grounds of an abandoned site are being used as a
 Nursery School.15  These sites have not been investigated and
 cleaned up. How many more potentially hazardous sites are being
 used  improperly? Answers are needed as well as elimination of the
 hazards.
 RECOMMENDED ACTIONS
   The  nation's approach  to  existing coal gasification  sites  has
 varied from well-planned at St. Louis Park, Minn, to an emergency
 response at Stroudsburg, Pa. There needs to  be a better, more
 deliberate approach.
   The mere presence of contamination may be taken as proof that
 a hazard exists. When investigating these sites, a balance must be
 sought between the legitimate concerns for public health and en-
 vironmental safety  and the need to bring land back into an ap-
 propriate and productive use.
   A  recommended method to deal with these serious problems at
 former sites containing coal tar residues follows:
 Phase I—Discovery and Evaluation
 •Location of all sites
 •Collection of historical information
 •Description of current and planned activities at each site
 •Preliminary description of geology and hydrology
 •Evaluation of potential hazards and urgency for remedial action
  based upon current and planned usage
 Phase II—Testing and Remedial Design
 •Detailed testing and evaluation  to determine the extent and na-
  ture of contaminants in the soil, aquifers and nearby waters
 •Install monitoring wells to measure movement of contaminants
  with time
 •Establish the potential hazards and identify options for remedial
  actions
 •Set  criteria for allowable concentrations of various contaminants
  in soil, aquifers and water
•Select the optimum scheme for remedial action at each site which
 will  meet the established criteria,  consistent with current and
 planned usage
Phase III—Cleanup and Monitoring
•Implement the selected remedial actions
•Establish  specific site use restrictions which should be mandated
 bylaw
•Provide for long-range monitoring to assure  that corrective ac-
 tions continue to meet specified  criteria and also that site restric-
 tions are not violated

CONCLUSIONS

   Many of the sites which remain to be evaluated will pose serious
enough hazards to health and the environment to rate them above
many on the current Superfund lists.  Coal tar residues  contain
known carcinogens and priority pollutants. Governmental agencies
need to take immediate action to deal with these sites in the most
expeditious manner to prevent further serious damage to people
and the environment.

REFERENCES
  1. Rhodes, E.G., "The History of  Coal Tar and Light Oil,  Bituminous
    Materials: Asphalts, Tars and Pitches,"  Coal  Tars and Pitches, 3,
    A.J. Hoiberg, ed., Robert E. Krieger, Publ. Co.,  Melbourne, FL,
    1979.
  2. Villaume, J.F., Lowe, P.C. and Unites, D.F., "Recovery of Coal
    Gasification Wastes: An Innovative Approach," Proc.  of 3rd Na-
    tional Symposium  on Aquifer Restoration and Groundwater Moni-
    toring,  Columbus, OH, May, 1983.
  3. Yazicigil, H.  and Sendlein, L.V.A., "Management  of Groundwater
    Contamination by Aromatic Hydrocarbons in the Aquifer Supplying
    Ames, Iowa," Groundwater, 19, 1981, 648-665.
  4. Thompson, S.N., Burgess, A.S.  and O'Dea, D., "Coal Tar Contam-
    ination & Cleanup, Plattsburgh, New York," Proc. National Con-
    ference on Management of  Uncontrolled Hazardous  Waste  Sites,
    Oct., 1983, Washington, D.C., 331-337.
  5. Hull,  M.F.  and  Schoenberg,  M.E., Preliminary  Evaluation of
    Groundwater Contamination by Coal Tar Derivatives, St. Louis Park
    area, Minn., USGC, open-file report #81-72, 1981.
  6. Hickock, E.A., Erdmann,  J.B., Simonett,  M.J., Boyer, G.W. and
    Johnson,  L.L.,   "Groundwater  Contamination  with  Creosote
    Wastes," presented at  the National Conference on Environmental
    Engineering,  Minneapolis,  Minn., sponsored by the Environmental
    Engineering Division of the Amer. Soc. of Civil Engineers,  1982.
  7. Ehrlich. G.G., Goerlitz, D.F., Godsy, E.M. and Hull, M.F., "De-
    gradation of  Phenolic Contaminants in Groundwater  by Aerobic
    Bacteria: St. Louis Park, Minn.," Ground Water, 20, 1982, 703-715.
  8. Marean, J.B., "Coal Tar: One Utility's Approach to Dealing with a
    Widespread Problem," presented at EEI Biologists Workshop, Al-
    buquerque, NM, May 1982.
  9. Grimshaw, T.W.  and Little,  W.M., Remedial Measures Plan for a
    Spill of Solvent Refined Coal Liquid at the SRC Pilot Plant, Ft. Lewis,
     Washington,  DOE/ET/10104-T10.  Report  by  Radian Corporation,
    Austin, TX, Aug. 1980.
 10. Personal communication with Mr. Sam Gianti, State of New Jersey
    Dept.  of Environmental Protection, Hazardous Site Mitigation Ad-
    ministration, Trenton, NJ.
 11. Air and Soil Sampling and Analysis, Recreational  Park in Belmar,
    N.J., report to Jersey Central Power and Light by EBASCO, Aug.
     1983.
 12. Wilson, D.C. and Stevens, C., Problems Arising from the Redevelop-
    ment of Gas Works and Similar Sites,  Report #HL81/3178 (CIO)
    Environmental and  Medical  Sciences  Division,  AERE Harwell
    Laboratories,  Oxfordshire,  England, Nov. 1981.
 13. Smith, M.A., Redevelopment of Contaminated Land: Gas Works
    Sites,  Interdepartmental Committee on the Redevelopment of Con-
    taminated Land (ICRCL),  UK  Dept. of the Environment, London
    England, 1983.
 14. Personal communication with an anonymous contact in Kentucky.
 15. Personal communication with  G.H.  Gockley, Pennsylvania  Power
    and Light Company, Allentown, PA.
                                                                                    SITE ASSESSMENT & DISCOVERY
                                                                                                                               15

-------
                     AN  UNDERGROUND TANK  SPILL  RISK
                                   ASSESSMENT PROGRAM

                                             DANIEL F. PREDPALL
                                          Woodward-Clyde Consultants
                                                Wayne, New Jersey
                                           WARREN ROGERS, Ph.D.
                                         Warren-Rogers Associates, Inc.
                                              Newport, Rhode Island
                                             ALAN LAMONT, Ph.D.
                                          Woodward-Clyde Consultants
                                             Walnut  Creek, California
INTRODUCTION

  In this paper, the authors suggest a methodology by which one
can  rapidly  and inexpensively  assess potential  risks from un-
protected underground steel tanks and use this information to im-
plement a cost-effective program of risk reduction. This method is
particularly useful when the budget for investigation is limited and
funds must be allocated carefully. In this case, the immediate prob-
lem is to identify those tanks which should be investigated in detail.
One wishes to concentrate the investigation budget on those tanks
which potentially present  the greatest risks and avoid investigating
those tanks which present little or no  risk.
  The authors propose using a screening procedure which evaluates
each tank on the basis of inexpensive and readily available informa-
tion. For a relatively small cost, a company can identify those tanks
with high, intermediate and low risks. It can then allocate the
budget for detailed investigations to those tanks posing the highest
risk.

How Large is the Problem?

  There is mounting evidence that a major source of groundwater
contamination in the United States is from leaking  underground
storage tanks containing  fuel oils, gasoline, various solvents and
other toxic materials. Recent estimates put the total number of in-
stalled tanks at about two  million. Studies have indicated that up to
20% of these tanks may be leaking. Assuming a leak rate of only 1
gal/day/tank, tens of millions of gallons per year of toxic materials
could be discharging into the environment. These numbers indicate
why leaking underground tanks are  rapidly becoming a major
pollution control issue receiving high priority  attention by  the
USEPA, the states and local governments. Thirty-three states have
underground tank  regulations  either on the  books or under
legislative consideration.

Why Evaluate Tank Risk?

  Many firms that use  significant numbers of unprotected under-
ground steel tanks are at risk. A tank spill at a facility can involve a
number of potential problems. The spill, if it has migrated off-site,
can result in injury or damages to persons or property. The tank
regulatory  environment is changing almost daily,  and continued
regulatory  compliance requires frequent communications  with
regulatory  agencies. Litigation and site remediation  costs can be
high. Media coverage in these instances can have a detrimental ef-
fect on business. Political  and civic involvement could be extensive
as well.
  Therefore, there is a strong impetus to  be knowledgeable about
the present risks and act early to reduce the risks where practicable.
                                                      GENERAL APPROACH
                                                      Implementation

                                                        A spill risk assessment program must be an ongoing effort with
                                                      periodic review of all tanks. In any one period (e.g., fiscal year),
                                                      there are three steps that one should take:
                                                      •Screen the tanks to determine which will  be investigated  this
                                                       period
                                                      •Investigate those tanks,  identifying those needing  replacement
                                                       and/or site remediation
                                                      •Carry out the appropriate actions
                                                        The last two steps should be planned individually for each tank
                                                      or site, depending on its particular circumstance. The first step is
                                                      discussed in this paper. That initial phase gets the process started by
                                                      quickly surveying all the tanks and their respective environments.
                                                      Since all  tanks are included, the process must use a  uniform ap-
                                                      proach to all tanks. Also, since all tanks are included, it must be
                                                      relatively inexpensive. For a company with several hundred tanks,
                                                      even spending $1,000 per tank  would be prohibited for this initial
                                                      screening phase. Consequently, the  screening procedure must be
                                                      based on relatively inexpensive and easily obtained information.
                                                        During each  period, the process is repeated and  the tanks/sites
                                                      having  the highest risks are identified and dealt with, thus substan-
                                                      tially lowering their risks. The risks presented by the entire popula-
                                                      tion  of  tanks eventually  will  be  reduced  and  controlled at
                                                      reasonable levels.

                                                      Risk Defined

                                                        The word "risk" has several commonly used technical and collo-
                                                      quial definitions. In the context of this paper, risk can be described
                                                      as the probability of an event occurring that is associated with an
                                                      adverse consequence during a  stated period of time.  An adverse
                                                      consequence is one that  produces harm to a human population or
                                                      damage to the environment. The risk level or severity  is dependent
                                                      upon the event probability and  the magnitude of the adverse conse-
                                                      quence.
                                                        Risk  assessment, then, is the general term for the study of deci-
                                                      sions subject to uncertain consequences. Risk estimation is the
                                                      calculation of risk level,  and risk evaluation is the process of deter-
                                                      mining the significance of the estimated risks and planning actions
                                                      to deal with those risks.
                                                        The  two components of risk, the potential for  adverse conse-
                                                      quences and the likelihood of an initiating event, can be shown in
                                                      the form of a matrix (Fig. 1). The upper right corner of the matrix
                                                      represents a situation involving a high probability for occurrence of
                                                      an initiating event (e.g., a lank leak) combined with a high poten-
16
SITE ASSESSMENT & DISCOVERY

-------
   13
   o*
   UJ
                            INITIATING EVENT
                                (CAUSE)
                          	».
                         INCREASING LIKELIHOOD
                             Figure 1
                        Components of Risk


tial for adverse consequences (e.g., a large population exposed)
resulting in high risk level. The  opposite corner  of the  matrix
represents low risk level.
Underground Tank Spill Risk

  The risk associated with underground tanks can be expressed in
terms of the matrix shown in Figure 2. In this figure, the potential
adverse consequence is environmental or public health hazard and
the initiating event is a tank leak. Situations where tank leakage
probability is low  and the potential environmental hazard is also
low have  an acceptable risk; no  action is needed at  this time.
However,  the risk  is unacceptable where tank leakage probability
and potential environmental or public health hazard are high and
some action is needed to reduce the overall risk. Risk with regard to
underground storage tanks, then, can be expressed as the combina-
tion of tank leak probability and the potential for environmental or
public health hazard.
        S
        g
        UJ
       o§
       (C M
       NO
       <0
       Zu
       UJ W

       IS
INTERMEDIATE HI1K
                        TANK LEAKAGE PROBABILITY
                          (INITIATING EVENTS)
                                       LEAK RISK ESTIMATION

                                         A method is required to provide an estimate of future risk of
                                       tank leakage. Tank tightness testing will only provide information
                                       on the condition of a tank at the time of the test. Computation of
                                       the Soil Aggressiveness Value (SAV), a numerical procedure that
                                       relies on basic soil characteristics such as pH, moisture, resistivity
                                       and sulphides, and tank age can explain only a small porportion of
                                       data variability.
                                         In an attempt to develop a better predictive tool for underground
                                       tank leaks, the American Petroleum Institute (API) engaged one of
                                       the authors to  perform a statistical analysis of the occurrence of
                                       corrosion  failures  in unprotected  underground tanks  and  to
                                       develop a mathematical model of the process which  leads to tank
                                       failure.
                                         The first step consisted of data collection from API member
                                       companies. The result of this survey was a chart showing the fre-
                                       quency of leaks by tank age (Fig. 3). Leaks were reported as early as
                                       2 years after installation and as late as 45 years. Thus, it was clear
                                       that tank age was a very poor predictor of tank failure.
                                         An unprotected steel tank installed in clean backfill without im-
                                       purities and not  subject  to abrasion during installation  will ex-
                                       perience external corrosion which is essentially uniform over its sur-
                                       face. Such a tank will, for all practical purposes, last the useful life
                                       of the accompanying facility. Previous studies have shown that ap-
                                       proximately 23% of all installations corrode evenly.
                                         In the remaining 77% of the tanks observed, one or more localiz-
                                       ed anodes were established on the tank surface during installations.
                                       Typically,  these anodes consisted  of impurities in the  backfill
                                       material, abrasion of mill scale, failure to remove  shoring, etc.
                                       Under these conditions, the corrosion leading to perforation pro-
                                       ceeds  at   a  pace  determined  by  the  chemical  and  physical
                                       characteristics of the  backfill. The relevant  variables governing
                                       electrochemical reaction in the tank environment are electrical
                                       resistivity of the soil, pH (acidity), moisture content, sulphide con-
                                       tent of the soil and size of tank.
                                         The approach taken in the study was based on the hypothesis
                                       that, in  the presence of electrochemical corrosion, age  to leak
                                       should be proportional to the probability that a leak has  started,
                                       while 30% had a high leak risk—greater than 78% (Fig. 4).  What
                                       the profile will  look like in 2 years if no leak detection program is
                                       implemented is shown in Figure 5. At that time, 41% of the tanks
                                       will have a high leak probability.
                                           "1
r
--Jin™
1
11
In
lirunlLr,n,,n
                            Figure 2
                  Underground Tank Risk Matrix
                                                                   Figure 3
                                                              Tank Age to Leak

                                        An example of a cumulative probability distribution over the age
                                      at which  a leak begins calculated from the mean age relation
                                      presented earlier is shown  in  Figure  6.  Based  upon the  soil
                                      characteristics of this sample, the mean age to leak is about 12
                                      years. A similar plot for a soil sample taken from a different facility
                                      is found in Figure 7. In this case, the mean age to leak is 23 years.
                                      Thus, soil conditions can have a significant impact on the expected
                                      life of a tank.
                                                                                   SITE ASSESSMENT & DISCOVERY
                                                                                                                              17

-------
  The soil corrosion test data can be used to determine if a tank is
uniformly corroding by employing a tank tightness test on the high
part of the probability curve. One test will reduce the probability of
leaking from its maximum of 77% (note that 23% of all tanks cor-
rode evenly and will not leak) down to 50%. A second tightness test
a year later will reduce the probability to 20%, and a third test will
reduce the probability of point corrosion to a negligible 1%.
  In summary, the soil corrosion test is a quick, reliable and cost-
effective technique for estimating the risk of tank leakage. Using
this technique, one can determine the probability that the tank will
begin to leak before the end of the coming period (i.e.,  before the
next cycle of evaluation and investigation begins).
                                                        vertically into the water table and then travels horizontally to some
                                                        point of contact without being detected and removed or contained.
                                                        The point of contact could be a water well, a body of surface water,
                                                        a basement or utility conduit. The number of people affected will
                                                        depend upon  the population density in the adjacent  area.
                                                          Thus, the site hazard evaluation function is basically a product
                                                        of two terms: the first term is a surrogate for the probability that
                                                        the leak would reach a point of contact undetected, and the second
                                                        term is the population density. It is assumed that the surrogate for
                                                        the probability that the leak reaches a point of contact undetected
                                                        be cased on the time it would take to migrate to the nearest point of
                                                        contact (i.e.,  sensitive receptor). The calculation of this time is
  <

 o
 IT
24



22


20


18


16


14


12


10
u
IX
                .000   .002    .02    21    78    .97    99    1 00
                            Figure 4
                         Company Profile

SITE HAZARD ESTIMATION

   The hazard of a given site makes up the second component of
risk—the potential for adverse consequences. Site hazard, in turn,
has two  elements:  (1) environmental hazard and (2) public  health
hazard.
   Environmental hazards include aquatic habitat impairment, con-
tamination of  productive wetlands  and  loss  of habitat. Public
health hazards  include contamination of surface and groundwater
supplies used for drinking water, inhalation of gases or fumes that
have migrated through the soil to confined areas such as basements
and explosion hazards from those fumes.
   The site hazard  evaluation methodology  described below relies
on easily obtainable information and is, therefore, quickly and in-
expensively implemented.
   Since it is not feasible to do a complete risk analysis for each tank
at this stage,  the authors use an evaluation  function that  is a sur-
rogate for the expected number  of people  affected by a leak, given
that a leak occurs.  It is a surrogate in the  sense that it includes the
major  factors that describe  the risk  and  should be strongly cor-
related to the expected risk that  would be determined by a  rigorous
and thorough analysis.
   With respect to public health hazard, it is assumed that the public
will only be affected by a  leak in a tank if the material first seeps
                                                                                      Figure 5
                                                                Company Profile Probability of Leak Two Years from Now

                                                         based upon the vertical distance to the water table and the horizon-
                                                         tal distance to the nearest point of contact.
                                                           The time for  the material to migrate vertically to a water table
                                                         can be computed as a function of the depth to the water table, the
                                                         permeability of the soil and  the  net  precipitation. The  time to
                                                         migrate  horizontally to the nearest  point  of contact is  simply a
                                                         function of the  distance.
                                                           Finally,  these times must be converted into a surrogate for the
                                                         chance that the leak will actually reach the nearest point of contact
                                                         undetected. If the  time were  short (i.e., one month), the chance
                                                         would be very large, with a probability near unity. If the time were
                                                         long (i.e., 24 months), the chance would be low. It is likely that the
                                                         leak would be detected. This  leads to a function shaped like that
                                                         shown in Figure 8.

                                                           The form of the site  hazard evaluation function is:

                                                                 R = C (Tv + Th) x  P                             (1)
                                                         where:

                                                             TV   = Vertical time of travel to a water table
                                                             Th   = Horizontal time of travel to nearest point of contact
                                                             P    = Population density within a given radius of the tank(s)
                                                         C(time)   = as defined in Figure 8
18
         SITE ASSESSMENT & DISCOVERY

-------
      1.00

      090


      n BO

      0.70


      0.60




      040


      0 X


      020

      0 '0

      0 00
MEAN AGE AT 12


/I
1
I I 1 1 1 1 1 1 1 1 11
                             Figure 6
          Cumulative Distribution Over Tank Age at First Leak
                              ACTUAL AGE
                            Figure 7
         Cumulative Distribution Over Tank Age at First Leak
RISK EVALUATION
  Once the tank leak probability and site risks have been evaluated,
they are multiplied to give the overall risk. It is also instructive to
plot each tank on a figure such as Figure 2, showing tank leak pro-
bability on one axis and site risk on the other.
  At  this point, there  is an  urge  to define what  regions of  the
matrix represent acceptable versus unacceptable risks in economic
terms. However, since the level of acceptable risk is greatly depen-

                                                                                             TIME TO REACH NEAREST
                                                                                           POINT OF CONTACT (MONTHS)
                                                                                                Figure 8
                                                                                       Probability of Leak Detection
                                                     dent upon the risk attitudes of the decision maker, judgments on
                                                     acceptability cannot be made quantitatively. In comparing the risks
                                                     to the benefits derived, a weighting factor is necessary to take into
                                                     account risk perception when translating  the risk into economic
                                                     terms.
                                                       One  can conveniently divide the risks shown in Figure 2 into
                                                     three broad zones. The facilities with clear problems, such as a high
                                                     tank  leak probability and/or  high potential  environmental or
                                                     public health hazard,  fall into the highest  risk category. Facilities
                                                     with low tank leak probability and corresponding low potential site
                                                     hazards fall into the lowest risk category.  All other facilities then
                                                     fall into the intermediate category (facilities for which insufficient
                                                     information is known for category selection also go into the in-
                                                     termediate level).
                                                       This procedure provides two useful results. First, it gives a priori-
                                                     ty  ranking of the tanks.  This priority can  be used to identify
                                                     tanks/sites to be investigated during the coming period. The tanks
                                                     posing the greatest total risk should be investigated first, up to the
                                                     budget  limitation for tank investigations in this period. The second
                                                     useful result is some insight into the source  of the risk for each
                                                     tank. When the tank is plotted in Figure 2,  one can see not only
                                                     how high the risk is, but also whether it is due to risks of a leak in
                                                     the tank or due to high expected consequences given there is a leak.
                                                     This  information is essential for planning the next stage of in-
                                                     vestigation on the tanks.
                                                       Because of the potentially high investigation, cleanup and litiga-
                                                     tion costs associated with some underground tank spills, it is felt
                                                     that the value of the information obtained in the risk assessment
                                                     procedure described here justifies the expense of investigation and
                                                     repairs. Taking the long view of the underground tank problem will
                                                     result in better resource allocation and lower costs overall.
                                                                                    SITE ASSESSMENT & DISCOVERY
                                                                                                                  19

-------
   DETECTION OF GROUNDWATER CONTAMINATION BY
  SHALLOW SOIL GAS SAMPLING IN THE VADOSE ZONE
                         THEORY AND APPLICATIONS

                                         ERIC G. LAPPALA
                                     Harding Lawson Associates
                                          Novato, California
                                      GLENN M. THOMPSON
                                    Tracer Research Corporation
                                           Tucson, Arizona
INTRODUCTION

  Groundwater contamination by volatile organic compounds is a
widespread problem resulting from the disposal and spillage of
solvents used in a wide variety of industrial processes. Several of
these chemicals, particularly the halogenated organics, have action
limits set at levels significantly less than 100 /ig/1. These limits have
been, or are now being, established by state and federal regulatory
agencies (Table 1). For example, an action level of 5 and 10/ig/l for
trichloroethylene (TCE) has been used as a guideline for requiring
aquifer restoration in California. Similar limits are being used in
Arizona.
  To date,  the discovery and  definition of the area! extent of
volatile organics  in  groundwater has relied upon the installation
and sampling of borings and monitoring wells. While samples from
such installations provide the  best  method to  quantitatively
                                               measure the presence or absence of volatiles in groundwater, they
                                               have the following disadvantages:
                                               •They are expensive to install, develop and sample
                                               •The level of contamination is not generally known until after the
                                                well has been installed, developed and sampled and the results
                                                returned from the chemical laboratory
                                               •Many wells are often required to adequately define the extent of
                                                the plume to evaluate potential aquifer restoration methods

                                                 In this paper, the authors present theory and field data that
                                               demonstrate the applicability of soil gas sampling and measurement
                                               of volatile compounds as a method  of detecting the presence of
                                               these volatile compounds in groundwater. The method can provide
                                               a cost-effective means to supplement data collected by conven-
                                               tional means.
                                                  Table 1
                   Some Common Solvents and Associated Compounds (bat have been Detected in the Groundwaler

                                                                     Limits for



Compound
carton tetrachlorlde
cnlorobenzene
p-dlchlorobenzene
1,2,4 trlcnlorooeozene
ethyl benzene
1,1 dlcnloroet/iene
1,2 dlcnloroetnane
1,1,1 trlchloroetnane
1,1,2 trlcnloroetnane
1,1,2,2 tetracnloroethane
cnloroe thane
tr 1 chlorome thane
dlchlorme thane
chlorome thane
dlcftloro dlfluoromethene
trlchloro fluoromethane
trlchloroeUiylene
cnloroe thylene
1,1 dlchloroethlene
tetrachloroe thylene
acetone
Isopropyl alcohol
toluene
«ylenes
cyclohexenone
methyl ethyl ketone



Formula
CCl,
CfiH5Cl
C6H<,C12
C<^3C13
W/S
CH3CHC12
C1CM2CH2C1
CH3CC13
CHC12CH2C1
C12C:CC12
W1
CHC12
C1CH2C1
CH3C1
CCl2f2
CCl/
CHC1:CC12
CH2:CHC1
CCH:OC1
CC12:CC12
CHjCOOHj
CHjHjCHj
C6H5CH3
C6H4(CH3)2
CfiHlOU
Cr^coc^
Vapor
Pressure Boiling
NWHg Point
• 20-25 «C *C
99 77
133
174
213
136
5fl
63
75
17 114
121
12
61.
40
-24
-30
24
87
2300 -14
48
121
56
33 82
111
135
136 156
80


Aqueous
Solubility
No
No
No
No
siiom
Sllfpt
No
No
No
No
No
Sll9lt
SllSpt
Slight
No

SI l^i t
SllOlt
Slight
No
Yes
Yes
No
No
Slltfit
Yes


Specific
Gravity
i.sa
1.10
1.45
1.46
0.87
1.17
1.26
1.32
1.44
1.62
0.92
1.48
1.33
0.92

1.49
1.46
0.91
1.20
1.62
0.79
0.79
0.87
0.86
0.95
0.80
Long Ten Proposed
Exposure Kc-


EP* HAS EnforceaoleUJPrlorlty
44 0/1 Standards
0.4 4.5 •
72




.71 .95 •
3800 1000 *


0.10 (2)
150 -
-
1600 -
- -
2.8 4.5 •
1-2 - •
70 -
0.9 •


- 340
- 670


Poll ul»n t
X
X
X
X
X
X
X






X
X
X
X
.

.
_
X
_


        Data Sources: Chemical Rubber Company Handbook of Chemistry and Physics, USEPA Multimedia Environmental Goals For Environmental Assessment, Hawlcy. 1981
                 (I) USEPA Region 9, personal communication, "indicate! compound for which standards are proposed.   (2) As total Irihalomelhanes
20
SCREENING TECHNIQUES

-------
PREVIOUS STUDIES

  Gas phase transport through porous media has been described by
numerous investigators. Penman,' cited by Glauccum and others,3
described the movement of acetone and carbon disulfide. Albert-
sen,1 in Swallow et a/.," measured changes in the carbon dioxide
content of soil gas and used these as indicators of metabolic activity
over a plume of biodegradable pollutants in an underlying aquifer.
Weeks et a/.14 used the downward migration of two atmospheric
fluorocarbons  (CC13  F,  [Freon-11] and  CC12 F2,  [Freon-12])
through a thick vadose zone to evaluate the parameters describing
such movement as a diffusive process. Diffusion of fluorocarbons
through the unsaturated zone was measured by Thompson and
Kraemer12 in an investigation of gaseous diffusion potential relative
to low level radioactive waste disposal.
  Recently, Glauccum et al.3 used shallow soil gas measurements of
benzene to define a contamination plume containing both volatile
organics  and  electrically conductive compounds  in a shallow
aquifer. Soil gas measurements were made with a portable organic
vapor analyzer (OVA) and were all above the 1.0 ppm  detection
range of that instrument.
  Swallow and Gschwend"  obtained  data from  a controlled
laboratory experiment to show that trichloroethylene moves up-
ward from the water table into the capillary fringe as  shown in
Figure  1. They also presented data indicating measurable soil gas
concentrations of benzene, toluene and TCE from depths of 25 and
50 cm above the water table found at a depth of about 120 cm.
  In a recent theoretical  study,  Jury  et al.6  described  the
mechanisms responsible for the movement  of both  liquid and
volatile organic pesticides through the vadose zone.
  To the authors' knowledge, field data demonstrates measurable
quantities of volatile  compounds in soil gases found  above con-
taminated aquifers at depths greater than a few meters.
          17-
                                37.5              750

                        CONCENTRATION
                           Figure 1
             Trichloroethylene Concentration vs. Depth
                 (from Swallow and Gschwend")
THEORY
  For volatile organics present in groundwater to be detected in the
vadose zone, they must move  upward  from the saturated zone
through the capillary fringe and then to  the point from which the
soil gas samples are taken. In this section, the mechanisms by which
such movement occurs are discussed. This discussion shows that
water table fluctuations enhance and in some cases may be required
to provide significant upward movement of volatile contaminants.
The relationship of the saturated zone, capillary fringe and vadose
zone and the dominant transport mechanisms in each are shown in
Figure 2.
           Grcynd Surface
                                             Ground Surface
      HORIZONTAL DISTANCE
                          Figure 2
      Schematic Illustration of the Mechanisms Responsible for
  Contaminant Transport in and above a Contaminated Aquifer,
     the Relationship Between Liquid Saturation and Depth for
                    an Equilibrium Profile
and
  Neither retardation of volatile organics by adsorption on soils
nor biodegradation are considered in this paper. For the volatile
compounds measured during this study, these mechanisms are con-
sidered less significant than those that  will be discussed.'

Transport in Saturated Sediments
  Transport of contaminants  through the zone that is  fully
saturated with  liquid and in which the liquid is under positive
hydrostatic pressure occurs by advection, hydrodynamic dispersion
and molecular diffusion.
  Advection, or transport with  moving groundwater at the mean
groundwater velocity, is usually the dominant transport mechanism
in, the saturated zone. For relatively coarse-grained aquifers, in the
absence of significant recharge  or groundwater extraction, such
transport is  usually predominantly horizontal. Advective flux in
any dimension is described by Darcy's law corrected for porosity: V
= (K/Ne) AH,  where K is the hydraulic conductivity tensor, Ne is
the effective porosity, AH is the gradient of hydraulic head and A is
the interstitial groundwater velocity. The advective flux of a con-
taminant at concentration C is given by Qa = VC.
  Hydrodynamic dispersion describes  transport caused by varia-
tions in the hydraulic conductivity of  the porous  media in direc-
tions parallel and transverse to the direction of mean groundwater
flow.  Dispersion is  commonly  expressed as a linear function of
velocity in the direction of flow:  Dx = «XV, where Dx is the disper-
sion coefficient in direction x, V is the mean groundwater velocity
and ax is a characteristic  length or dispersivity  in  direction x.
Dispersive flux  of a contaminant is described by the product of a
dispersion coefficient and the concentration gradient as Qd = Dx?
C/aXj.
  Swallow and  Gschwend" attribute the  vertical movement of
volatile organic solutes above a uniformly contaminated aquifer to
the vertical  component of transverse hydrodynamic dispersion
associated  with horizontal groundwater  flow.  Transverse
hydrodynamic dispersion probably accounts for significant vertical
movement only when significant vertical flow components in the
                                                                                          SCREENING TECHNIQUES      21

-------
small scale  velocity  field are present  in an aquifer. Transverse
dispersivities are typically one-half to two orders of magnitude less
than the longitudinal coefficients.
  Crane and Gardner showed in 1961 that for a uniform sandstone
the ratio between the transverse and longitudinal dispersion coeffi-
cients ranges from 0.10 at velocities less than 0.01  m/day to less
than 0.01 at velocities of 10 m/day. For the horizontal flow experi-
ment in a uniform sand conducted by Swallow and Gschwend," a
vertical transverse characteristic length or dispersivity of 0.0033 m
was reported. Groundwater velocities were not given in this study
to enable computation of the dispersion coefficients.
  When flow is predominantly horizontal, a commonly found field
situation, vertical transverse dispersion may be less significant than
found in the laboratory study of Swallow and  Gschwend. Under
field conditions, water  table fluctuations  may provide  a more
plausible mechanism for transport of contaminants through the
capillary fringe and into the vadose zone.
  Molecular diffusion  describes transport caused by a spatial  gra-
dient of the concentration of a solute and is given by a generalized
Pick's first  law: Qd  =  Nc Db dC/dXj, where  Db is the bulk liquid
diffusion coefficient and Ne is the effective porosity. As described
in  subsequent sections, diffusive flux  through  saturated zones is
generally small compared to other mechanisms.
  Liquid-solid partitioning, or adsorption, is important for some
solutes.  However, the  adsorption  or retardation potential for the
volatile compounds under conditions of full water saturation is low
and is not considered further.

Transport in the Capillary Fringe

  By the definition used in this paper, the capillary fringe is  that
zone above the water table which is fully saturated with water but
in which the liquid water is held under negative pressure or tension.
This zone, also referred to by some authors as the tension saturated
zone, is illustrated in Figure 2. With a number of discrete pore sizes
present, the thickness of this zone is equivalent to the pressure head
required to  empty the  largest pores. This pressure head is also re-
ferred to by some authors as the air entry or bubbling pressure2.
Since natural sediments typically  have areally  varying  pore  size
distributions, the top of the capillary fringe may not be a planar
surface as depicted in Figure 1.
  Under steady flow conditions, transport  through the capillary
fringe may  occur by the same mechanisms  as described  for the
saturated zone. If the water table does not fluctuate and the flow of
contaminated groundwater is predominantly horizontal,  the only
mechanisms by which contaminants can move across the capillary
fringe are transverse hydrodynamic dispersion and molecular diffu-
sion. For commonly encountered horizontal groundwater velocities
(0.05 to 2 m/day), the vertical flux due to transverse hydrodynamic
dispersion is proportional to the transverse dispersion coefficient.
This coefficient is of the order of 1 x  10-3 mVday for a velocity
of 0.1 m/day and OT of 0.01 m. Diffusive flux is proportional to the
liquid  diffusion coefficient  which is of the  order  of 1  x 10 ~5
mVday. Consequently,  contaminant  flux  rates   through  the
capillary fringe caused by these two mechanisms would be very
slow for all  but materials having a very high transverse dispersion
coefficient  caused by  anomalous  vertical heterogeneities in the
aquifer materials.
  A fluctuating water table above a contaminated aquifer may pro-
vide a more rapid mechanism by which  volatile organics may move
into the vadose zone. Figure 3 shows a simple case of a water table
rising rapidly from position (1) to position (2). This rise pushes un-
contaminated water in  the capillary fringe upward into the vadose
zone. When the water  table falls,  as shown in  position (3), con-
taminated water will be retained in the vadose zone and throughout
the  capillary fringe. Hysteresis in the relationship between pressure
head and water  content enhances  the  retention of  contaminated
water in the vadose zone under these conditions of water table fluc-
tuation. This enhancement occurs because, at a given tension, more
water is  retained in the pores as the water table is  lowered than
enters the pores as the water table rises. This  hysteresis in the
                                                                                    Figure 3
                                                          Schematic Illustration Showing Contamination of the Vadose Zone and
                                                              Capillary Fringe by a Rising (2) then Falling (3) Water Table
                                                         pressure head-water  content  relationship is usually  more  pro-
                                                         nounced for coarse-grained soils near saturation than  for  fine-
                                                         grained soils (Hillel, 1971).

                                                         Transport in the Vadose Zone

                                                           The presence of volatile contaminants in and above the capillary
                                                         fringe provides the opportunity for their upward transport at  rates
                                                         several orders of magnitude greater than those under conditions of
                                                         full saturation.
                                                           Under conditions of  no significant recharge and no redistribu-
                                                         tion of soil moisture, the two dominant mechanisms of transport in
                                                         the vadose zone are gas-liquid partitioning and gaseous diffusion.
                                                           Contaminant flux caused by gaseous diffusion is described by
                                                         Pick's first law applied to a gas filled pore space: Qg  = Dg dCa/dz,
                                                         with Dg = 0a TDab, where 0a =  the air filled pore space, t  = tor-
                                                         tuosity and Dab = the diffusion coefficient of gas a into gas b. It is
                                                         assumed that the "a" is the volatile contaminant, and that "b" is
                                                         the soil gas.
                                                           The gas phase  diffusion coefficient is about 104 to 103 times as
                                                         large as the liquid phase diffusion coefficient.6'14 Weeks  el al."
                                                         computed gas phase diffusion coefficients for the fiuorocarbons
                                                         CC13F and CC12F2 of 0.78 and 0.86 mVday based on an  empirical
                                                         equation developed by  Slattery and  Bird." The equation used by
                                                         Weeks et al. is reproduced here because few measured values for
                                                         Dab are available  in the literature for many of the compounds given
                                                         in Table 1 .
Dab  A/P (Pa
(Ta  Tb)5/12
                                                                                                        [T/(Ta Tb)]B  (1)
                                                          where Pa, Pb  = critical pressure for gases a and b, in atmosphere;
                                                          T«. Tb = critical temperatures for gases a and b, degrees K; p =
                                                          ambient atmospheric pressure, atmospheres; Ma, Mb = molecular
                                                          weight of gases a and b in g/mole; A =  2.745 X 10 -4; B =  1.823;
                                                          T = ambient  temperature, in degrees K.
                                                            Jury et al. give empirical equations for the gas phase diffusion
                                                          coefficient through a porous media:
                                                                                                                    (2)
                                                                                                                     (3)
                                                          These authors concluded that both  the  aqueous  and gas self-
                                                          diffusion coefficients (Dc and Dab) were relatively constant at the
                                                          values of 4.3  x  10-5 and  0.43 mVday,  respectively,  for in-
                                                          termediate weight molecular compounds such as most pesticides.
                                                            As shown in Figure 3 and in Equation 2,  the gas diffusion coeffi-
                                                          cient is directly proportional to the air-filled porosity. Hence the
                                                          opportunity for upward diffusion increases as drier soils are found
                                                         and for the liquid diffusion coefficient:

                                                                 ,,,10/3,
22
SCREENING TECHNIQUES

-------
closer to the land surface under conditions of insignificant recharge
and redistribution of soil moisture.
  Gas liquid partitioning.  Once volatile contaminants are present
above the capillary fringe, they will tend to partition between the li-
quid and gas phases. Measured partitioning coefficients for several
volatile  organics detected  in field  samples during one of the
authors' studies are shown in Table 2.
  The larger the value of Kw, the more volatile the compound and
the larger the tendency to be present in the gas phase and to be
available for transport by gaseous diffusion in the vadose zone.
FIELD METHODOLOGY

  The authors have made measurements of soil gas concentrations
of volatile organics at 14 sites in a variety of geologic and climatic
environments during the past 18 months. The  field methodology
developed and used at these sites is presented in this section.
  Soil gas samples are obtained from shallow depths by driving a
hollow,  perforated metal probe to the desired depth. Vertical pro-
files to  depths  of less than 3 m are determined  by driving and
sampling to successive depths. Vertical profiles deeper than 3 m are
obtained by  driving the soil gas sampling probe ahead of the bot-
tom of a hollow-stem auger that has been advanced to just above
the desired sampling depth. Soil gas is pumped from the sampling
location at a rate of 2 to 3 1/min by a peristaltic pump. The soil gas
sample is collected in a glass syringe by insertion of a clean needle
into the sample line. The gas sample is  then directly injected into
the gas chromatograph in the field.
                                     Table 2
                         Gas-Liquid Partitioning Coefficients for
                          Some Common Volatile Compounds
                     Kw = concentration in air/concentration in water
Compound
F-ll
F-12
F-ll 3
CH2 C12
TCA
TCE
PCE
Kw
0.7
0.3
0.25
0.37
0.5
0.33
0.43
              Soil gas and water concentrations of the volatile organic com-
            pounds discussed in this section were made  in the  field using a
            Varian gas chromatograph equipped with electron capture (EC)
            and photoionization (PI) detectors. The  procedure incorporates
            proprietary methodology that enables chromatographic separation
            of the aliphatic  compounds  reported  in  this study in less than
            5 min. The separation of the aromatic compounds is usually ac-
            complished in less than 10 min. Detection limits are between 0.001
            and 0.01 /tg/1 f°r gas  samples and between 0.1 and 1.0 /tg/1 for
            water samples. All samples are replicated,  and reported results are
            the arithmetic mean of at least 2 replicates.
                                                         Table 3
              Summary of Field Studies of Groundwater Contamination Using Soil Gas Sampling and Analysis by HLA and TRC
Site Location or
Descript ion
Depth to
Ground
Mater
Moisture
Content
of Vadose
Zone
Clays in
Vadose
Zone?
Volatile
Organics
Detected
                 California State
                    Superfund  Site
25
                                                                      Freon-113, TCE,
                                          moderate        yes      TCft,PCE
                 South Valley  CERCLft Site             moderate
                    Olbuquerque,  NM             15       to  wet
                 Federal CERCLft Site,
                     Tucson,  ftrizona            188       dry to
                                                           moderate

                 Electronics Firm,
                     Southwestern U.S.           80         dry

                 County Landfills,                       dry to
                     firizona                      100      moderate

                 New  Electronics Plant,
                     Northern California        60      moderate

                 fierospace  Plant,  Northern
                     California  CERCLfl Site    50         dry
                          yes      Benzene,TCE,
                                    Toluene,TCO
                          yes
                                    TCE,TCP,PCE,
                                    CC13H
                            no      TCE, TCfi
                          yes      TCfl,TCE,PCE
                            no      TCft, TCE, DCft
                                    DCE
Electronics  Firm,
    Northern  California

Chemical  Storage  and
Transfer  Facility,  Calif.
                                                           moderate
                                                            to  wet

                                                           moderate
                                                            to  wet
                           yes
                                    TCE,Freon-113
                                    TCfl,TCE,DCE,
                           yes      PCE,Benzene,
                                    Toluene
                 Service Station,
                    Northern California
25
        moderate
                                                                             yes       Benzene,
                                                                                       Toluene
                                                                                     SCREENING TECHNIQUES
                                                                   23

-------
  Many of the studies have included the sampling of water from
existing monitoring wells and analyzing the water samples in the
field with the methodology described above. Water samples are all
taken with standard sampling protocols for volatile organic com-
pounds. Duplicates of water samples are sent to independent, cer-
tified laboratories for quality control testing.
  Field quality control for gas sampling involves the following:
•Flushing of the sampling probe, tubing and pump by drawing at-
 mospheric air through the system until concentrations of the com-
 pounds of interest are at atmospheric levels as determined by
 periodic samples of the ambient atmosphere at the site
•Frequent analysis  of blanks and known standards in the field

APPLICATIONS
   To date, Harding Lawson Associates and Tracer Research Cor-
poration have used the soil gas method to detect and define plumes
of volatile organic  contaminants in groundwater at 14 sites. These
sites have provided a data base for evaluating the applicability and
limitations of the method under a wide range of subsurface condi-
tions. All field studies have been conducted where the vadose zone
comprises non-indurated sediments. The types of sites and subsur-
 face conditions at 10 of these sites are given in Table 3. This section
describes the results of studies at three of these sites in more detail.

California State Superfund Site,
 Northern California

   This site was chosen to verify the method because several plumes
of volatile organics have been delineated by the installation and
sampling of monitoring wells. A series of aquifers at the site ranges
in depth from a few to several hundred meters and consists primari-
ly of fine sands to coarse sands and gravels.  The zones between the
 aquifers  are  typically clays to silty  clays.  Elevated levels  of
 Freon-113, 1,1,1-trichloroethane (TCA), TCE and other volatile
organics have been found in the shallow aquifer. This contamina-
 tion resulted from leaks of underground tanks and pipes used for
 the storage  and transmission of  these solvents. The following
sampling activities  were conducted at this site:
 •A  vertical profile over an area where the shallow aquifer was
  known to be  uncontaminated
 •Two vertical  profiles over areas where the shallow aquifer was
  known to be contaminated
 •A  horizontal transect across a plume that had been well defined
  by monitoring wells
 •An areal survey to determine the extent of volatile compounds in
  the shallow aquifer
At all of these sites, the water table was between 25 and 33 ft (7.5 to
 10.5 m) below the  land surface.
   Vertical profile over an uncontaminated aquifer.  This site was
located up-gradient of a  known source of contamination. The
depth of water at  this site is 24 ft (7.2 m). Table 4 contains the
results of the analyses performed for methylene chloride, F-113,
TCA, TCE and PCE.  The trace levels of the compounds found are
lower than the detection level of the laboratory analysis method for

                            Table 4
             Chemical Data for the Uncontaminated Site
                   (all concentrations are in /ig/l)
Sample
Air above
ground (1)
Soil gas 10 ft
(7.5 m) (1)
Soil gas 25 ft
(10.5 m) (2)
0.005 ± 0.005

Water (I)
(field meas.)
Water (ind.
lab analysis)
CHjCIJ F-113 TCA

O.I 0.004 0.003

0.02 0.04 0.003


0.09 ± 0.01 0.01 ± 0.01 0.001 ± 0


«I.O) 0.3 0.2

ND
TCE

(40.001)

0.001


0.001 ±
0.001

(<0.l)

ND
PCE

0.002

0.05





O.I

ND
                                                         the compounds in water. The trace levels indicated in Table 4 may
                                                         be due  to a lower level  of decontamination procedures used for
                                                         sampling equipment than used for the remaining sites. In contrast
                                                         to the sites discussed below, no vertical trends or patterns are evi-
                                                         dent in  the data.
                                                            Vertical profiles over  a  known contaminated aquifer. Soil gas
                                                         profiles were sampled at two sites over a plume of volatile organic
                                                         compounds that had been mapped using conventional drilling and
                                                         sampling methods. Data collected at one site are found in Table 5.
                                                         The increases in concentration of TCA, TCE and PCE with depth
                                                         at a second site  are shown in Figure 4. The data shown in 4a sug-
                                                         gest:
                                                         •The relative proportion of volatile compounds in the soil gas
                                                          phase  roughly  corresponds to  predictions based upon the gas
                                                           liquid partitioning coefficients given in Table 3
                                                         •The soil  gas concentrations  are not  in  equilibrium  with  the
                                                          groundwater as would be predicted based on the partitioning
                                                          coefficients alone
                                                            Data in Table 5 show a similar decrease in concentration with
                                                         distance above the water table.  However,  with the exception of
                                                         TCA, they  also  show a decrease from the soil gas immediately
                                                         above the water table to the concentration below the water table.
                                                         The distribution of compounds at this site is not an obvious func-
                                                         tion of their aqueous solubility as appears  to be the case for the
                                                         data shown in Figure 4. This may imply separate incidents of the in-
                                                         troduction of contaminants into  the subsurface.
                                                            Ul V.n.cjl Profiln Of V«
                                                                                                   l«f. SwiU CUr» County. C*liforfiu
                                                                                   COMCthTHAtiOMWl
                                                                     161 V.nlal Piofiln O( Volltlb 0*5.0.0 Abo«t A Conunituml Aquifn. Tuaon. Arirau
                                                                            Abb rev iat ions d re:
                                                                                            TCA
                                                                                            PCE
                                                                                            F"U
                                                                                            F-11J
                                                                                            TCE
                                                                                          1.1,1 tnchloro«lh*no;
                                                                                          pcrchloroethylcnej
                                                                                          crichlorofluoromethantt
                                                                                          trifluorotrichloroethAne
                                                                                          trichlorovthylene
                                                                                              Figure 4
                                                                    Vertical Profiles of Volatile Organic Concentrations in Soil Oas above
                                                                    Contaminated Aquifers in Tucson, Arizona, and Santa Clara County,
                                                                                             California
 24
SCREENING TECHNIQUES

-------
                           Table 5
                    Chemical Data for Site 2
                  (all concentrations are in
Sample

Air above
 ground (1)

Soil gas 5 ft
 (1.5m) (4)

Soil gas 15 ft
 (4.5 m) (2)
Soil gas 20 ft
 (6 m) (4)
Water (1) (field
 meas.)
Water (ind.
 lab analysis)
 CH2C12    F-113     TCA      TCE     PCE


  0.1       0.2     (<0.001)   (C0.001)   (C0.001)

1.5 ± 0.8   3.5 ± 0.1  0.14 ± 0.08  0.01 ± 0  0.45 ± 0.2


170 ± 23    71 ± 6     2 ± 1   0.60 ± 0.14  5.0 ± 0

190 ±100   100 ± 32   4.0 ± 1.8   0.9 ±0.1    6 ± 6


 29 ± 5   65 ±  13   120 ± 29   0.6 ± 0.3  0.1 ± 0 .1

                    70      100     0.50
  Transect across a known plume. To evaluate the soil gas sampl-
ing and  field analysis  as the methods for detecting plumes of
volatile contaminants, a series of samples was taken and analyzed
along a line transect across a known plume of TCE and TCA. The
depth to water at this  site is 30 ft  (9 m). The  results of soil gas
sampling compared to  concentrations  found in monitoring wells
completed in the shallow aquifer are shown in Figure 5. Soil gases
at this site were obtained at depths of from 2Vi to 3 ft (0.75 to 0.90
m). The correlation between data obtained by the two methods as
shown in Figure 5 is sufficient to use soil gas sampling and analysis
as a semi-quantitative technique at  this site for detecting ground-
water contamination.
  Arealsurvey. Soil gas samples were taken from depths of 2 to 3 ft
(0.6 to 0.9 m) in an area where the limits of groundwater containing
elevated levels of volatile organics were poorly known. This survey
comprised samples taken from 54 soundings as shown on Figure 6.
The sampling and field analysis by GC  for F-113 was completed in
under three days. Concentrations shown on Figure 6 are averages
of two replicate measurements.
  In addition to a rapid survey to define the extent of F-113, this
survey was able to delineate a source in the upper part of the study
area that was not defined prior to field measurements. The extent
of contamination mapped by the soil gas samples was subsequently
confirmed  by obtaining  water  samples from borings and wells at
several locations as shown on Figure 6. The degree of correlation
between volatile organic concentrations in the groundwater and the
soil gas is high. The real time field GC analytical capability com-
bined with  the rapid means of obtaining soil gas samples resulted in
a significant  cost savings over  conventional drilling and sampling
methods.
  The subsurface variability of the  parameters that control gas
phase diffusion is such that a theoretical prediction of concentra-
tions in groundwater from concentrations in  soil gas is  probably
not feasible.  However, site-specific calibration of the method can
provide at least a semiquantitative prediction of groundwater con-
centrations. The correlation between soil gas and groundwater con-
centrations of F-113  at  the site of the  areal survey is shown  in
Figure 7. Although scatter exists in the data, the correlation over
four orders of magnitude is reasonably good. This correlation can
be established by obtaining groundwater samples using conven-
tional means at a limited number of control points.
CERCLA Site, Southern Arizona

  This site is characterized by an arid climate. The depth  to the
shallowest  permanent groundwater in the area where the soil gas
survey was  conducted is approximately 100 ft (30 m). The aquifer at
this site is contaminated with TCE, PCE and 1,2 Dichloroethylene
(DCE). A vertical profile is reported  for this site. The concentra-
tions of PCE, TCE, CC13H and Freon-11  as a function of depth are
shown in Figure 4b. The presence of a 25-ft (7.5 m) thick sequence
of clay in the vadose zone is significant. This clay is apparently dry
enough or  contains sufficient macropores to allow significant up-
ward gaseous diffusion of the compounds found in the ground-
water.  Upgradient from the area of the survey, the clay serves as a
perching layer. Perched conditions were not encountered at the site
of the profile.
                    • Soil Gas from 3.5 feet deep
                    O Groundwater

                      •PREVIOUS STUDY BY HLA
                                                       DEPTH TO GRCUfCmTER~3O FT
                                                    400
                DISTANCE FROM REFERENCE POINT (feet)
                           Figure 5
               Soil-Gas Transect Across TCE Plume
                         l AREAS or MO XCESS CUC TO
                         • PARKING, BUILDINGS, OR OTHER .^L*
                         -V STIWCTUKS JJT--JTT-. :-
                                                                                            EXPLANATION

                                                                                S^'lo  Soil 9" "mplinfl location and F- 113 concentration in ug/l


                                                                               f>Ł™  F" '13 conc'nt'«i»n in groundwater in ug/l
                                                                               Figure 6
                                                    Concentrations of F-113 in Soil Gas Above a Shallow Aquifer  and in
                                                      Groundwater from the Shallow Aquifer at an Industrial Complex


                                                                            SCREENING TECHNIQUES       25

-------
  All compounds found in the soil gas except Freon-11 apparently
have their source in the ground water beneath the site. Freon-II
concentrations  show a  gradient that indicates a surface or at-
mospheric source. The data obtained by Weeks el at." showed a
similar gradient which was attributed to the downward diffusion of
Freon-11  from  elevated levels in the atmosphere.  Since these
elevated levels are present essentially worldwide, it is reasonable to
assume that the atmosphere at the site is the source of Freon-11 and
that the groundwater is  acting as a sink for this compound.

Electronics Manufacturing Plant,
Southwestern United States
  This site is underlain  by groundwater at a depth of about 80 ft
(24 m) and is contaminated with TCE, TCA, 1,1  DCE, Methylene
Chloride and PCE. The vadose zone at the site comprises fine sands
to coarse gravels with little silt and clay. The climate at the site is
arid, and the moisture content of  the vadose zone is low. This soil
gas survey was implemented to evaluate the extent of any off-site
contamination  to determine if off-site monitoring wells and/or
aquifer restoration might  be required.  Additional detail in  the
plume configuration was also needed to enable optimal location of
on-site extraction wells for aquifer restoration. The study was con-
ducted in two phases:
•A survey of the property boundaries and on-site sources
•An off-site transect perpendicular to the direction of the ground-
  water flow
   Contours of soil gas concentrations of TCE and TCA are com-
pared  in Figures 8 and 9 to contours based upon water  samples
from nine monitoring wells  that  are screened in the top of the
aquifer.  The  agreement  between  the  general plume patterns
mapped  by both methods is apparent from these figures. The soil
gas survey provided additional detail  in areas where no monitoring
well data were available. In particular, evidence of suspected up-
gradient contamination from another suspected  source was not
                     Sample Silt Piln

                   GroundwiCtr  Soil Gu
                            F-113 Concrntratloru 0-J/D

                             Groundwilrr  Soil Gu
                     074 2
                     A 13
                     744
                     0743
                     074-6
                     074 5
                     074 7
                     B8
                     SG3
                     SG39
                     SG 11
                     SG 14
                     SG-47
                     SG-44
                     SG-49
                     SG32
1 4
580
eooo
so
t 6
32
34
I
1 I
600
6900
250
48
1000
18
52
     10.000
       100-
                                                    10.000
                           Figure 7
Relationship Between F-113 Concentrations in Soil Gas and Groundwater.
         Sample Numbers Refer to Sites Shown in Figure 6.
                                                         found, and the zone having the highest concentrations along the
                                                         southwestern boundary was delineated.
                                                           The extent of off-site groundwater contamination was evaluated
                                                         by conducting a transect across the plume about  250 m  down-
                                                         gradient from the site boundary. Both TCA and TCE are present at
                                                         significantly lower  concentrations than  those  found on-site (Fig.
                                                         10). No wells or borings have been installed to confirm this pattern.
                                                           To evaluate the  reproducability of the method,  repeated sam-
                                                         pling was done at five sites on successive days. The second sample
                                                         was taken from the same depth and within 5 ft (1.5 m) of the first
                                                         of each set. The results of this repeated sampling comparison are
                                                         shown in Table 6. Students t test analysis of the paired  data shows
                                                         no significant difference between sample means taken on successive
                                                         days. The t test was done on the logarithms of the raw data because
                                                         of the  wide range  in values. The  F-ratio test showed  the sample
                                                         variances to be the same at the 0.001 significance level, justifying
                                                         the use of the Students t test.
                                                           Six of the on-site soil gas samples were taken in the immediate
                                                         vicinity, within 10  ft (3 m) of existing  monitoring wells.  Water
                                                         samples taken  from these wells using dedicated Teflon and PVC
                                                         bladder pumps  were  analyzed  in the  field with   the  same
                                                         methodology used for the soil gas samples. The correlation between
                                                         the soil gas and groundwater concentrations is shown in Figure 11.
                                                         The regression was performed on the common logarithm  of the
                                                         values because of their wide range.
CONCLUSIONS
  Theoretical considerations of the mechanisms by which volatiles
may reach the shallow sampling depths from contaminated ground-
water include the following:
•The principal vertical transport mechanisms under conditions of
 no recharge or water table fluctuation are: transverse hydrody-
 namic dispersion through the saturated zone and the capillary
 fringe, liquid phase diffusion through the capillary fringe and
 gas phase diffusion through the vadose zone.
•Water table fluctuations coupled with hysteresis in the water con-
 tent-pressure  head relationship can  greatly accelerate the intro-
 duction of contaminants into the vadose zone.  A significant water
 level rise followed by a decline will  be more effective in provid-
 ing this  introduction of contaminants above the  capillary fringe
 than other sequences of  water level fluctuations.  Water level
 fluctuations may, in fact, be required in many situations to pro-
 vide significant gas concentrations at the base  of the vadose zone
 to detect soil  gas contaminants at shallow depths.
•The introduction of volatiles into the capillary  fringe and the
 vadose zone by water level fluctuations in a contaminated aquifer
 may  result  in a  significant degree of subsurface contamination
 that may impact the time required  for aquifer  restoration. In
 some  instances,  aquifer restoration  by groundwater extraction
 may need to be enhanced by  flushing of these zones by artificial
 recharge as part of the overall remedial action for contaminated
 aquifers.
  Field data obtained in this study demonstrate the following:
•The presence  of volatile organic compounds in groundwater may
 be detected by analyzing soil gas samples from depths as shallow
 as 1 m for water levels at about  10 m and from as shallow as 3 m
 for water levels as deep as 30 m.
•Vertical   profiling of the  concentrations of volatile  organics
 found in soil gases provides a vertical concentration gradient
 which may be used  to infer whether aquifer contamination or  a
 surface source is responsible for the observed concentration pro-
 file.
•When field analyses of the volatile compounds are made,  the
 shallow  soil gas sampling method conducted along transects can
 provide  a rapid real time assessment of the extent of subsurface
 contamination. As such, the method may be used to augment
 conventional drilling and sampling methods to more economically
 provide  additional data on the extent of contamination.
 26
SCREENING TECHNIQUES

-------
         LOG OF   TCE   IN GROUND WATER
LOG  OF   TCE   IN  SOIL GAS
     LEGEND
   • Monitoring Well
   A Soil Gas Probe
/-»3.r Log of TCE Concentration in Groundwater
     Log of TCE Concentration in Soil Gas
         100
                      200 meters
                                                       Figure 8
   Comparison of TCE Concentration Distribution as Determined by Groundwater and Soil Gas Samples, Southwestern US Study
          LOG  OF   TCA   IN  GROUND  WATER
LOG OF   TCA   IN SOIL GAS
      LEGEND
    •  Monitoring Well
   A  Soil Gas Probe
  J t  Log of TCA Concentration in Groundwater
  J r  Log of TCA Concentration in Soil Gas
         100
                      200 meters
                                                      Figure 9
   Comparison of TCA Concentration Distribution as Determined by Groundwater and Soil Gas Samples, Southwestern US study
                                                                                      SCREENING TECHNIQUES       27

-------
                                                                                                   Table 6
                                                                              Results of Repeated Sampling, Southwestern US Study
                            Figure 10
    Transect Down-gradient from Property, Southwestern US Study
               	o
              TCA IWI • 7 • • O.ft' ISGI
                    i* -oar
                    0123

                        LOG Of SOIL GAS CONCENTRATION < SC I
                            Figure 11
  Relationship between TCA and TCE Concentrations in Soil Gas and
             Groundwater for the Southwestern US Study
•Because of  the  approximately exponential decrease in concen-
 trations from the water table to the land surface, soil gas detec-
 tion limits for volatile organics may need to be less than 0.01 jig/1
 to detect moderate to low levels of contamination  from water
 tables at depths greater than 10 m.
•At sites where tight, very wet to saturated clays are present, or
 where the contaminated aquifer lies beneath  a clean aquifer, the
 soil gas method cannot be used effectively to detect groundwater
 contamination.
Sampling
Polol
soio
SO 100
SCI 10
SG50
SG45
TCA
5/8/84
1.8
2.9
2.9
315
220
5/9/84
1.9
3.2
2.7
200
172
TCE
5/8/84
4.0
.85
3.6
675
240
S/9/S4
4.1
.99
3.3
360
200
                                                             REFERENCES

                                                              1.  Albertsen, M. and Matthess, G., "Ground air measurements as a tool
                                                                 for mapping and evaluating organic groundwater pollution zones,"
                                                                 International Symposium on Groundwater Pollution by O.I Hydro-
                                                                 carbons, Prague, 1978, 235-251.
                                                              2.  Brooks,  R.A.  and  Corey,  A.T.,  Hydrauiic properties  of porous
                                                                 media. Colorado State University Hydrology Paper 3,  1964.
                                                              3.  Glaccum. R., Noel,  M. and  McMillan, L., "Correlation of geo-
                                                                 physical  and organic  vapor  analyzer data over a conductive plume
                                                                 containing volatile organics," Proc. of the Third National Symposium
                                                                 on Aquifer Restoration and Groundwater Monitoring. National Water
                                                                 Well Association, 1983, 421-427.
                                                              4.  Hawley,  G.G., The Condensed Chemical Dictionary,  Van Noslrand
                                                                 Reinhold Company, New York, NY, 1981.
                                                              5.  Hillel, D.. So/7 and  Water.  Physical principles and processes. Aca-
                                                                 demic Press, New York. NY, 1971.
                                                              6.  Jury, W.A., Spencer. W.F. and  Fanner. W.J., "Behavior assess-
                                                                 ment model for trace organics in soil: I. Model description," /. Envir-
                                                                 on. Quality. 12,  1983, 558-564.
                                                              7.  Penman, H.L., "Gas  and vapor measurements in soil, 1, the diffus-
                                                                 ion of vapors through porous solids." J. Agric. Set., 30,1940,437-462.
                                                              8.  Penning ton, D., "Retardation  factors in aquifer decontamination of
                                                                 organics: in Aquifer  Restoration  and Groundwater Rehabilitation,
                                                                 Proc. of the Second  National Symposium on Aquifer Restoration
                                                                 and Groundwater Monitoring, National Water  Well Association,
                                                                 1982, 1-5.
                                                              9.  Pickens, J.F.  and .Grisak, G.E.. "Scale dependent dispersion in a
                                                                 stratified aquifer; Water Resources Research. 17, 1981, 1191-1211.
                                                             10.  Slattery, J.C. and Bird, R.B.,  "Calculation of the diffusion coeffici-
                                                                 ents of  dilute gases  and of the self-diffusion coefficient of dense
                                                                 gases, AtChEJ, 4. 1958. 137-142.
                                                             11.  Swallow, J.A.,  Gschwend,  P.M.,  "Volatilization of organic  com-
                                                                 pounds  from  unconfined aquifers, in Proc. of the  Third National
                                                                 Symposium on Aquifer Restoration and Groundwater Monitoring,
                                                                 National Water Well Association, 1983. 327-333.
                                                             12.  Thompson,  G.M. and  Kraemer,  O.K.,  "In situ measurement of
                                                                 fluorocarbon diffusion rates in unsaturated media," Annual Report
                                                                 to U.S.  Geological Survey, Contract 14-08-001-20430, 1981, 17.
                                                             13.  Thompson, G.M., Demonstration of soil-gas sampling as a tool to
                                                                 aid in defining the distribution of subsurface contamination by vola-
                                                                 tile  organic compounds. Consulting report to  Harding Lawson
                                                                 Associates, Novato, CA, 1983.
                                                             14.  Weeks, E.P., Earp, D.E. and Thompson, G.M., "Use of atmospheric
                                                                 fluorocarbons F-ll  and F-12 to determine the diffusion  parameters
                                                                 of the unsaturated zone in the southern High  Plains of Texas,"
                                                                 Water Resources Research,  18, 1982, 1365-1378.
28
SCREENING TECHNIQUES

-------
                      QUALITY  CONTROL ATTRIBUTES  OF
                             PROCESS  ANALYTICAL  DATA

                                         PAUL H. FRIEDMAN, PhD.
                                            Viar and Company, Inc.
                                              Alexandria, Virginia
                                               DUANE GEUDER
                                    U.S. Environmental Protection Agency
                                                Washington, D.C.
INTRODUCTION

  The National Contract Laboratory Program (CLP) is a nation-
wide network of laboratories under contract to the USEPA. The
Program was originally designed to supply routine chemical analy-
sis services to the Hazardous Waste Site Investigation program.
These laboratories work under firm, fixed-price  controls  using
standardized analytical methods, sample handling procedures and
data reporting protocols.1 The CLP also has the capability and re-
sources to supply these same analytical services to other programs
and agencies.
  The CLP provides the USEPA with analytical data on which to
base work assessments, to  institute remedial action or to initiate
enforcement procedures in  order to contain or properly dispose of
identified hazardous waste.  The Program is  structured with a
strong orientation toward supporting enforcement activities. Pro-
tocols and methodologies are USEPA-approved to provide data of
documented  quality using  analytical quality control (QC) pro-
cedures and a system of document control.
  The primary objective of the CLP is to provide routine, high
volume analysis of samples collected  from hazardous waste sites.
Using a single  national program for this effort: (1) ensures that
all samples are analyzed according to uniform and consistent pro-
tocols which is a vital requirement for enforcement actions, and
(2) achieves low analysis costs through the economies of scale. All
CLP analyses are performed by private laboratories of proven abil-
ity which have won competitive contract awards. Continued high
quality laboratory performance is assured through ongoing evalua-
tions conducted by the Environmental Monitoring and Support
Laboratory/Las Vegas (EMSL/LV). EMSL is also  responsible for
developing all methods, standards and protocols used by the con-
tractor laboratories. Final data review and evaluation is conducted
by the CLP support staff with assistance from EMSL.

Analytical Methodologies
  Standardized analytical protocols are used for all routine work
carried out within the CLP activities.1 In addition to the methods
of analysis, the protocols specify the quality control precedures
and documentation requirements. A generalized schematic of the
analytical protocols is given in Figure 1. The need for uniform and
consistent protocols is in juxtaposition to the almost infinite variety
of samples that are encountered in the Program.


INTERPRETATION OF DATA QUALITY
  The scope of the data requirements is imposed by the various in-
vestigation and remedial programs which generate the samples and
apply the data to litigation and engineering purposes. Data used in
these programs must be known and of documentable quality.
Objectives of Data Evaluation

•Reduce the probability of "bad data" not being identified
•Increase the amount of usable data by resolving technical prob-
 lems
•Create and use a system  of data quality indicators to guide the
 data users
•Assist in the application of data to particular policy questions
 given required quality parameters
•Evaluate new methods of analysis in terms of efficacy and appli-
 cability
•Evaluate laboratory method performance
•Evaluate the adequacy of methods and laboratories in the deter-
 mination of new compounds or in new matrices
•Assist in the application of data  to particular field problems,
 given required measurement and performance parameters
                        Figure 1
                    Analytical Protocols

  An analytical result, like a tangible product, goes through a man-
ufacturing process. Like a manufactured item, the analytical result
has identifiable attributes which can monitor the quality of the re-
sult. The evaluation of data quality occurs within the context of the
requirements placed upon the data. The estimation of quality is a
function of the information content of the data as they apply to
their intended  use. This conforms more to the classical definition
of product quality as the appropriate level of quality for the appli-
cation or function of the product.
                                                                                   SCREENING TECHNIQUES       29

-------
  The analytical process may be treated as a semi-continuous pro-
duction process in which the component processes contributing to
the quality of the overall results may be separated. Simple statisti-
cal techniques are used.2 These may be derived from the monitor-
ing, evaluation and qualifications of batch process manufacturing.
The evaluation of quality  requires knowledge of the behavior of
the controlling variables or  functions which correlate with  those
variables in the analytical process.
  Precision, accuracy and  recovery may be calculated for  analyti-
cal determinations if some of the following assumptions are in-
corporated into the model structure supporting these calculations:
•The  sample population is homogenous by virtue of the fact that
 the matrix  is in the  same category; i.e., all well  water  or river
 water.
•Compounds spiked into each sample are spiked at approximately
 the same level for a particular compound.

Components of Analytical QC
   The samples taken from a hazardous waste site are very different
from  samples  taken from  industrial processes, clinical/biological
sources or nearly every other type of system where sampling is a
means of obtaining information. The sample is less uniform with
respect to matrix as well as analyte. In spite of these very important
differences,  adequate statistical procedures can be employed to de-
termine the  precision and accuracy of  the  data  and to establish
realistic performance limits.

Surrogate Compounds
   Standard  laboratory analytical quality control procedures1  are
used  to monitor variables affecting the analytical process. The
analytical QC sample types and the categories of quality attribute
data  that are  derived from  these samples are listed in Figure 2.
Laboratories also employ  special techniques suited to hazardous
waste site environmental analysis. These techniques include the use
of surrogate compounds spiked into each sample. These procedures
help to derive information about the precision and accuracy of the
analytical function.
   The QC protocols for the analysis of hazardous waste site sam-
ples employ isotopically enriched organic compounds as analytical
tracers. These compounds  are isotopically  different  from their
natural abundance analogues, or they are the fluorinated analogues
of chlorinated pollutants.  The surrogate compounds, by the ana-
lytical fraction which they  monitor, are  listed in Table 1. The pro-
cedure differs from isotope dilution in which an isotope analogue
(radioactive  or stable) is determined and the results for the analyte
are normalized to  the recovery of the  isotopically enriched ana-
logue.
   Surrogate compounds are added to  the appropriate analytical
fraction to monitor the analytical procedure. Results obtained for
the surrogate compounds do not monitor individual compound re-
sults except  in the instance  where the  surrogate is an isomer  or
analogue of a particular analyte. To characterize the analytical
data,  the percent recovery of the surrogate compound is calcu-
lated  and normal statistical  attributes of the data are calculated.
The statistical  attributes calculated and the arithmetic formula for
each attribute  are given in Table 2. These attributes may  then  be
used to quantitatively define the analytical quality of the data de-
rived  from samples taken from a particular site and provide a
sound analytical basis  for  the  interpretation  of the  physical
phenomena indicated by comparative analysis of these data.
  Surrogate  compounds have several advantages relative to using
the split samples and spiked samples required in each study:
•Each sample contains  QC information on accuracy and  pre-
 cision. Statistically valid information can be generated much more
 rapidly than with the matrix spiked and duplicate samples.
•Surrogates  are totally synthetic  compounds; i.e., they  are  not
 found in the environment and interferences are minimized.
•Surrogates are introduced into each sample at the analyzing labor-
 atory. Analytical results for surrogates are totally independent of
 field  sampling procedures.
                                                                              >/
                                                                              
-------
                           Table 2
               Statistical Attributes of Sample Data
WITHIN BATCH PRECISION
BATCH TO BATCh PRECISION
                                        OTHER DEFINITIONS

                                       L  NUMBER OF ANALYTICAL
                                          BATCHES

                                       11.- NUMBER OF SAMPLES  IN
                                        1 THE iTH BATCH

                                       N  TOTAL NUMBER OF
                                          SAMPLES

                                       S.= STANDARD DEVIATION
                                        1 OF THE iTH BATCH
                                          AVERAGE RECOVERY,  X.
                                       N.- NUMBER OF SAMPLES
                                        1  PER BATCH

                                       X   GRAND AVERAGE OF
                                           ALL BATCHES
   WHERE:
SX  =
                         Zl*f-K)
                                    = .z Y/,
                            L-
 STANDARD ERROR OF A SINGLE MEASUREMENT
                      I.
 Shewhart QC Charts
   Originally developed for control of production processes where
 large numbers of articles were being manufactured and inspected
 on a continuous  basis, control chart concepts have  been readily
 adapted to laboratory operations where the analyst produces com-
 paratively fewer results on an intermittent basis.  Inherent in the
 approach is the recognition of the basic assumption that combina-
 tions of random  and systematic variations exist in every method
 and within every  laboratory. The mathematical relationships and
 facsimile of control charts are shown in Figure 3. Industrial accep-
 tance of control charting and other statistical techniques has grown
 out of the basic theories and procedures of Shewhart.2

 Evaluation of Site Data
   Data are  evaluated  on a study-by-study basis. Each  study  is
 composed of samples from only one site. The CLP refers to each
 study as  a Case.  The general procedure is to: segregate informa-
 tion by analytical batch;  check off required information; calcu-
 late statistical parameters; and,  if enough information is available,
 calculate control  charts based on the data presented. Such charts
 are meaningful only to the degree that enough information is avail-
 able to  obtain  a reasonable  representation  of  the  required
 parameters and that the data included within the information can
 be assumed to be homogeneous.

 PROCEDURES
 Segregation of the Analytical Data
   The data are related to each other as a consistent sample matrix
 type; as being produced by a particular laboratory; and as a part of
 a particular analytical batch. The relationship by matrix and labor-
 atory are self-explanatory. The analytical batch consists of those
 samples  that are processed  simultaneously or in a continuous se-
 quence under conditions that associate the samples.  Samples pro-
 cessed simultaneously by the same personnel using the same batch
 of solvents and methodology can be considered an analytical batch.
Data Grouping
  Recoveries data for each Case were arranged in the following
groups:
•Volatiles
•Base/Neutral
•Acids
•Pesticides
  Within each chemical grouping, the data were further subdivided
by surrogate compound. The data for each surrogate compound
are further subdivided by the combination of matrix and extraction
data. Examples of distinct matrices include the following:
•Drinking Water (Finished)
•Well Water
•Surface Water (Standing Bodies)
•Surface Water (Flowing Bodies)
•Surface Water (Leachate and Runoff)
•Air
•Soil
•Sludge
•Drummed Materials

   This process allows greater homogeneity of the sample popula-
 tion and identifies the analytical batch.
   The hierarchy depicted in Figure 4, matrix 2, ultimately contains
 two VOA batches, two acid extractable batches and two base/neu-
 tral extractable batches. Comparison of analytical  performance
 and results  can only be within a particular matrix. Parallel reduc-
 tion of data can be accomplished on matrices 1 and 3, but the data
 must not be mixed for calculations.
   A matrix of the characteristic data by batch is created. This in-
 cludes average recovery and standard deviation of the average re-
 covery. The types of summary data calculated for each surrogate
 compound  from the characteristic batch data are included in Table
 2. The value of determining important or controlling variables is
 illustrated by the data set in Figure 5. When these symbolic data
 are sorted by date, it can be readily seen that a systematic differ-
 ence occurred for those samples that comprise Batch C. This group
 of samples  was more likely subjected to a systematic deviation than
 a randomly occurring one. Isolating systematic deviations serves to
 isolate uncharacteristic data and helps  identify operational prob-
 lems or relationships in the laboratory or in the field.
   The summary of data collected from multiple Cases analyzed by
 the same laboratory for the same matrix over a period of time may
 be combined to construct a control chart using the relationship and
 charts in Figure 3. Most single Cases do not have enough chron-
 ological history  to establish reliable control values. The control
 chart will become more useful as an information tool as more qual-
 ity control data are stored.

 Results
   The batch data and  summary data are presented  in TAbles 3A
 and 3B for a Case of samples recently evaluated. Results  for the
 precision and accuracy are in terms of  the percent recovery of
 surrogate compounds.  The control limits for the Shewhart charts
 are also calculated. Since this set of data covers only a very short
 period of time and a small number of batches, the  control limits
 are not representative of overall laboratory  performance. The re-
 sults obtained for the limits of the data as indicated in Table 3 may
 indicate that some of these data are unsuitable for the barest qual-
 itative determination at the concentration level represented by the
 surrogate while other results are quantitative. The accuracy,  pre-
 cision and  control indicators enable the data user to quantitatively
 evaluate the site data and determine the significance3 between and
 among field results.
   As can  be seen from the tables, the same  samples yield  very
 different results  for  the different  analyte categories. The quality
 obtainable for the analysis of samples is a function of many inter-
 acting variables which  may be monitored but not necessarily  con-
 trolled to  the desired  degree.  While it is possible  and desirable
 to indicate a minimally acceptable degree of data  quality or in-
                                                                                           SCREENING TECHNIQUES
                                                            31

-------
          u
         •0  •


         K


         >0


         10

         0
                   I    I    1   4   »   i

                     MTCl. I
    (I)
 M


 M


 70


 M
          40  —


          M  -


          M  -


          It


          0
 BCl,



 I


 Ul,

•ICl,
                                                             TX AVERAGE  Of THE RAN&E Of RECOVERY fQR THE DELICATE  SA»«»US


                                                             1 0^                       OP»>E« CONTROL LIMIT


                                                             "  * * *P«  *                ^>Pe« WARNIN6 LIMIT
                                                       AVIRACES


                                                         K  IS THE &RA»C AVtRAW Of THE MLANS Of BAT&CS


                                                         OCL  « K  •  A I                UP»>1» COMTROL  LIMIT


                                                         LCL  • X  -  A,I                LO-C* CONTROL  LIMIT
                                                                      r X  -   J/J A •
                                                                                                              LIMIT

                                                                                                       WA*NIN& LIMIT
                                                               FOR  n
                                                             J.267

                                                             l.MO
                     UIC» I
                                                             Figure 3
                                                   Construction of Control Charts
 formation content, that degree may not be achievable due to a lack
 of an adequate method as surely  as operational difficulties. The
 QC procedures embedded in the analytical process are present to
 ensure the data generator and the  data user the ability to discrim-
 inate between these complications.
   Three control charts generated from a volatile organic surrogate,
 a base/neutral surrogate and an acid surrogate are shown in Figure
 6. The control  charts were derived from several analytical batches
 of samples arising from a particular study from a hazardous waste
 site. The  amount  of data generated  by several analytical batches
 from one site is insufficient to make a generalization concerning
 the control posture of the laboratory for those analyses. Inspection
 of the control and warning ranges for the determination of these
 compounds  indicates a wide variation in the recoveries and ranges
 of results to be expected when the mandated  methods are used to
 determine constituent analytes. The width of the ranges determined
 for the base/neutral compounds indicates that the variation in re-
 covery may be unacceptably large while the average recovery of the
 acid compound may be unacceptably low.
   The quantitative description of  data quality allows the user to
 judge the appropriateness of data in terms of the information con-
 tent of the data and the relevance of the data to the objective of the
 data-gathering activity. The user requirements for the data should
 not dictate how well defined the data are, but how useful the data
 are with respect to the defined quality. Data appropriate to certain
                                                          legal proceedings may  only have to establish the presence of the
                                                          material in amounts significantly different from background or en-
                                                          vironmental levels. Remedial requirements may demand the estab-
                                                          lishment of significance between concentrations of analytes in ad-
                                                          joining samples.
                                                            The operational aspects and the degree to which the quality of
                                                          the data are defined are approximately the same in both legal and
                                                          remedial situations. The requirements in terms of the extent of in-
                                                          formation contained in the data are different. The differences in
                                                          information content needed by the user may indicate that different
                                                          ways of generating the  required data may be appropriate. This, in
                                                          turn, may  require different techniques of data evaluation.
                                                          DISCUSSION

                                                            Quantitative procedures provide a tool which enables a more ra-
                                                          tional  interpretation of field  data as a  function of  physical
                                                          phenomena  through an understanding of the limitations on the
                                                          data. These procedures, when consistently applied, also provide the
                                                          means  for defining the analytical process as it relates to  labora-
                                                          tories, methods,  matrices and the inter-relationships among these
                                                          variables.
                                                            Using surrogates as a determinant of laboratory quality  has the
                                                          advantage of obtaining results that are unbiased by field sampling
                                                          and environmental contamination. That is, the replicate precision
                                                          of results  is unaffected by the precision of sampling since surro-
32
SCREENING TECHNIQUES

-------


Ml 1
timcTiw



Mr 2
EITMCTIW
                                MI 3
                              MALTS1S
                           Figure 4
            Segregation of Data into Analytical Batches

gate compounds are added to samples in the laboratory before
analysis. The presence of the compound in the field sample negates
contributions to inaccuracy and imprecision as occur in the situa-
tion with spiking actual pollutants. Using these compounds also
negates the  extra costs and resources required  to analyze more
samples in duplicate as background  and spiked samples or accept
a lower level of confidence due to  gathering infrequent  and less
representative QC data. Finally, the  use of a consistent set of com-
pounds at  similar concentrations across  many  laboratories and
within a laboratory over a long period of time allows the gather-
ing of data to establish performance characteristics of any variable
or group  of variables which can  be isolated in the analytical
process.
  As a probe of analytical quality, surrogate compounds have cer-
tain drawbacks. The surrogate compound behavior must be corre-
lated to the  analytical behavior of the hazardous compounds and
may not be  direct indicators of analytical  quality as related to in-
                                                                            _l— BBIL
                                                                                         o
                                                                            ini
                                                                                                 x
                                                                                                 IMI
                                                                                        w
                                                                                         n.
                                                                                         IL
                                                                            mi
                                                                            A,
                                                                            A
                                                                            EL
                                                                                                  iijj
                                                                                     UiL
                                                                                      TTT
                                                                                              iiit
                                                                                                          ^
                                                                                                           till

                                                                                                          Wv
                                                                                                           III!
                                                                                              ^

                                                                                Figure 5
                                                                             Symbolic Data
                                                             Top in numerical sample order; bottom sorted by date.
            X - 107»    O-TOLUENE
i   »   »   »

    BATCH
                                 UCLJ
70.
0-

OS-PHENOL
X 2"l\
sv= u»
St = 5t
L
-~-^----
_^y__.
1 2 3 "I
BATCH
                                                                                 100 -c:—	—zz—zz———zz—.11
                                                                                \  50-
                                                                                       X   85*
                                                                                       S =  7\
                                                                                                J-FLUOROBIPHENVL
                                                            Figure 6
                                             Control Charts of the Ranges and the Averages
                                                                                         1   2   J

                                                                                            BATCH
                                                                                          SCREENING TECHNIQUES
                                                                                                                33

-------
                          Table 3A
                    Batch Statistical Data (%)
                                                                                    Table 3B
                                                                                Summary Data (7e)

SURROGATE COMPOUND
0,-lOLUtNt
Ob- 1, 7-DICHLORO-
t THANE
BROHOFLUOROBENJCNe
Dj-NHROBENIENE
7-fLUOROBIPMEWfl.
B10-PT«EM
0llt-f.-TERPNENVL
Oj-PHENOl
7-f LUOROPHENOL


DIBUTUCHi.O«END»JE
BA1C" 1
»( JIO OIV
11 1 5
1
" ! '
11 ! -
SO 17
?•. i)
Id M
77 ' It
1
1
71 »
ti, 10
1
II ' »
1
,, ' ,.
BAH" )
«} 4TD. OIV.
11 )
1
11 1 I
" 1 "
1
M 1 10
I
1
Ik It
11 ! 1)
1
11 | l»
i
1
11 1)
>>l Ik
1
1
It 17

t
1- i II
	 1 	
lAICn )
jij $10. Dlv
11 )
t
11 , »
10 7
,k ' ,1
1
II 11
100 ' II
,1 ' ,.
1
II l 1
11 , 1)
kk ' 7k
	 _| 	
1) ' )
1
 dividual analyte behavior. Compounds that are added to samples
 may not faithfully mimic  the matrix/compound interactions  of
 those analytes that are environmentally present. Knowledge of the
 correlation of the analytical behavior of surrogate compounds with
 that of the analytes in various matrices would be helpful in under-
 standing the results of analyses.
  The more QC information on which data evaluations are based,
 the greater the probability that these evaluations  are describing
 reality and not anomalies.  This points up the need to automate
 the collection and analysis of QC data.
  Applying statistical and manufacturing quality control to labor-
 atory analyses provides an unbiased  procedure for gauging the
 quality of analytical data and thereby establishing the information
 content of the analytical results. These statistical attributes of the
 data establish a system for defining the accuracy and precision of
 the analytical data and the control posture of the analytical labor-
V«*M,(«T.,.
0( - IOUOC
* in«M 	
MMZCAC
0,-xi ncmoaot
I-'IUOMBIPXHII.
0.--TMK
Du-p-n«P>
I
11
1)
»
n
III
It
»
*'


»t
V
*
t
IS
»
II
77
17
"


II
t,
•
1
11
II
70
II
10
71


71
'l
1
*
II
II
7»
71
>-
It


It
i
1
1
M
II
II
II
1
II



L/UCL,
It/11
11/101
I/Ill
ll/ltl
ti/i>r
H/ltl
u/ti
IS/11

tt/in

«.
ii
it
it
in
F?
Ill
M
"

n

                                                         atory with respect to these procedures and samples. The variable
                                                         nature of the sample/method interaction requires that the data
                                                         evaluator take the perspective of defining the information content
                                                         for the data user rather than enforcing a  more or less arbitrarily
                                                         chosen quality standard. The analytical result must have a level of
                                                         quality associated with it consistent with its uses. If this level cannot
                                                         be achieved consistently, other paths to the result must be chosen.

                                                         REFERENCES

                                                         I.  Federal Register, 44. No. 233. Dec. 3. 1979.69501-69540.
                                                         2.  Inhaven, S., Ed., Quality Assurance Practices for Health Laboratories,
                                                            Amer. Pub. Health Assoc., Washington. DC, 1978.
                                                         3.  Natrella, Mary Gibbons,  "Experimental Statistics."  NBS Handbook
                                                            #91, U.S. Government Printing Office, Washington, DC, 1983.

                                                         ACKNOWLEDGEMENTS

                                                           The authors wish to thank Ms. Doris Ling for assistance in pre-
                                                         paring material for the manuscript, Mr. William Eckel for his re-
                                                         view and comments, and Ms. Barbara Jean for manuscript editing
                                                         and preparation. This work was funded by  the  U.S. Environ-
                                                         mental Protection  Agency  Office of Emergency and  Remedial
                                                         Response, Washington, DC, under EPA Contract No. 68-01-6702.
34
SCREENING TECHNIQUES

-------
         CHARACTERIZATION  OF ORGANIC WASTES FOR
    EVALUATION  OF  REMEDIAL ACTION ALTERNATIVES

                                           GREGORY A. MOONEY
                                                  CH2M HILL
                                             Montgomery, Alabama
                                           RUSSELL W. HARTLEY
                                   U.S. Environmental Protection Agency
                                                  Dallas, Texas
INTRODUCTION

  Using conventional sampling and analytical techniques, investi-
gators very often identify only a fraction of constituents present
in complex organic  hazardous wastes. Such techniques as gas
chromatography/mass spectrometry (GC/MS) used to determine
organic priority pollutants are widely recognized as vital to hazard
assessment  and  do provide  an essential initial  characterization.
However, the following constraints and considerations must be
realized.
•Many high molecular weight organic compounds cannot be iden-
 tified  or quantified by GC/MS methods due to system limita-
 tions and sample behavioral properties. If a portion of the sample
 is not soluble in the extraction solvent or will not aspirate, for ex-
 ample, it will not reach the system at all.
•All GC/MS systems were not created equal. System capabilities
 differ widely.  The minimal "economy" models can often only
 identify and quantify listed priority pollutants. Research-grade
 systems—especially, used in conjunction  with gel permeation,
 capillary tubes and other ancillaries—provide much broader capa-
 bility.
•In a complex waste matrix,  detection limits of many constituents
 may be quite high, even though these constituents may represent
 significant hazard or may affect use of certain remedial action
 technologies.
•Analytical or administrative standard  procedures often limit the
 information provided in conjunction with analytical reports. Due
 to standards of identification or quantitation confidence limits,
 statements like "present but not quantified" or "present below
 quantitation confidence limit" appear in detailed laboratory re-
 port appendices. Also, contract limitations for analytical services
 often pre-establish the number of organic  constituent peaks be-
 yond listed priority pollutants to be identified. Thus, conclusions
 in site investigation reports regarding presence or absence of cer-
 tain constituents or classes of chemicals are valid only to the de-
 gree that the analytical and  reporting procedures are comprehen-
 sive.
•The Contract  Laboratory Program (CLP) is designed for bulk
 conventional analytical services, with some provisions for special
 analytical services (SAS).  However, some capabilities simply are
 not available through CLP or require significant additional effort
 by all parties to arrange. As a result, the tendency to simply select
 analytical parameters from the available standard menu generally
 prevails.
  Of critical importance in evaluation of remedial action alterna-
tives is the assessment of whether specific technologies for destruc-
tion, hazard reduction or isolation are applicable and feasible for
a specific site. Often, these technology assessments are based on
one or two episodes of conventional sampling and analyses with
inherent constraints described above. Further, sampling methods
must be considered regarding representative results for each media.
  Inevitably, the gap between analytical results  and  feasibility
assessments proves very broad, yet it must be bridged.  For feasi-
bility assessments, a reasonable prediction of how a given waste
material will react in many engineering situations is required.
Predictive ability is fairly good for specific chemicals and moder-
ate for a few waste  materials and mixtures that have  been pre-
viously studied or tested.  However, no experience  base exists for
many complex organic matrices resulting from  random disposal
practices typical of uncontrolled hazardous waste sites.
  Additional characterization and testing methods, beyond con-
ventional analytical approaches but not including field trials or
pilot tests, can be utilized to bridge the gap between investiga-
tion efforts and feasibility assessments with improved confidence.
Analytical or behavioral testing  of the actual waste material,
preferably in or close to  the physical state and  condition antici-
pated for handling, is often needed. Some of these methods, devel-
oped for actual site investigations and feasibility assessments, are
described in this presentation.
  It is important that a site-specific approach must be  developed
in each case. Use of multiple laboratories, including both private
and contract laboratories, may be necessary to obtain the desired
capabilities for a given site investigation. Further, information and
data obtained for feasibility purposes may not require the normal
degree of quality assurance/quality control and evidentiary pro-
cedures.
FIELD TECHNIQUES
Pit Profiling

  Many uncontrolled hazardous waste sites have resulted from the
disposal of hazardous materials in open pits (Fig. 1). These waste
pits can contain multiple layers including floating material, water,
oils, other organic layers and bottom sludges. In addition to depth
probing, the nature and variability of stratification must be deter-
mined. While sampling with depth  in various  locations in waste
pits is generally necessary, the number of samples and cost of
analyses—in addition to labor and  expense—to effectively deter-
mine stratification can be substantial.
  After a review of numerous probing methods, a technique for
measuring in situ viscosity was developed for a site with numerous
                                                                                    SCREENING TECHNIQUES
                                                       35

-------
pits containing complex  organic wastes. A viscosity profile as a
function of depth was performed at multiple locations in each pit
using a  modified  Nametre Model 7.006 C4P  vibrating sensor
(probe) viscometer powered by a 12V battery (see Fig. 2). All fac-
tory-supplied gaskets were replaced with Viton gaskets.  A 1.0 in.
diameter steel pipe attached to the viscometer served as a handle
and as a conduit for the  transducer cable, which was also encased
in Teflon.  This unit had a digital  readout and a maximum range
of 0.1 to 100,000 centipoise (cp). The viscometer, connected at the
bell housing to a steel winch cable, was raised and lowered using
a manual winch with position-lock features.
   Prior to each use, the viscometer was zeroed and the calibra-
tion checked.  At each profile location and depth, the reading was
allowed to stabilize prior to recording the data. After each use, the
immersed sections were solvent-cleaned and rinsed.
                          •SURFACE OIL/GREASE
  SURFACE WATER
                                                               I
                                             IHTERBEDOEO CRUST

                                            - - r-1^: '• •	
          ORGANICS LAYER

            sorrow SLUDGE
                             Figure 1
                      Typical Waste Pit Section
   TRANSDUCER CABLE
   EHCASEO IN TEFLON
 EX TENSION DEPTH
 PROBES
   PROTECTIVE BASKET
                                               VOL T BA TTERf

                                        POWER CONVCR1ER
                          SENSOR TIP PROBE
                             Figure 2
                       Viscometer Schematic
   Total pit depths were measured at viscosity profile locations and
a few additional locations, using either a separate depth probe rod
or the viscometer equipped with two bottom probe extensions.
   This method  of assessing  variability and stratification proved
to be very effective, since viscosity variation between layers was
substantial.  Viscosity is a parameter needed to evaluate almost all
handling and processing alternatives. An additional benefit of de-
termining the viscosity  of materials is that viscosity is proportional
to molecular weight for many  organic substances.  Hence, data
gathered through in situ viscosity profiling was used to select loca-
tions for sampling, thereby reducing the number of samples.
                                                            The organic  strata profiled also contained numerous volatile
                                                          organics which  would be released if samples were brought to the
                                                          surface in open containers. In previous sampling efforts at this
                                                          site, observations of high rates of release of volatile chemicals and
                                                          corresponding  apparent  viscosity  increases had been  reported.
                                                          Thus, in  situ viscosity measurements  provided valuable  informa-
                                                          tion on handling properties such  as  pumpability of the various
                                                          layers.

                                                          Conlained-Volatiles Sampling
                                                            In most sampling, few precautions are taken to contain volatile
                                                          chemicals. In pits, in situ absolute pressures may exceed  7 psig at
                                                          16 ft liquid depth. A reduction in pressure as the sample is raised
                                                          to the water surface causes a release of volatiles. This loss occurs
                                                          when  one uses traditional sampling  techniques such as column
                                                          liquid sampling (coliwasa), thieves, dredges and pumping. Similar
                                                          problems occur with monitor well sampling methods.
                                                            Additional releases of volatiles can occur in subsequent  handling
                                                          of samples.  Resulting  analyses can therefore  indicate a  volatiles
                                                          content substantially less than actually present.  Also, due to the
                                                          above considerations, laboratory analyses may yield results below
                                                          detection limits or  of  low confidence,  particularly in a  complex
                                                          waste matrix.
                                                            The content and nature of volatile organic compounds  can have
                                                          a critical  impact on remedial action feasibility assessment  and ulti-
                                                          mate  implementation. Organic  constituents have  been demon-
                                                          strated to limit use of many fixation/stabilization alternatives.
                                                            Volatile release could cause considerable damage to public health
                                                          and the  environment  through ambient air quality degradation,
                                                          particularly  if a release is not expected during remedial action.
                                                          Further,  many  organic chemicals are reactive in  the liquid and/or
                                                          vapor phases under certain conditions.
                                                            At  the same  site  where in situ viscosity profiles were obtained,
                                                          the release of volatile chemicals was a major concern.  In  prior
                                                          sampling efforts,  releases had been observed during sampling and
                                                          contained-volatiles sampling of organic liquids  in multiple pits was
                                                          deemed necessary.  To achieve  sampling objectives, crit1"-1'  re-
                                                          quirements of the sampling method and equipment were to:
                                                          •Collect and maintain the same at in situ pressures
                                                          •Contain the volatile compounds
                                                          •Prevent exposure to air, thereby reducing opportunities for oxida-
                                                           tion and other reactions
                                                          •Provide a  container  of  sufficient volume to  be cooled (iced),
                                                           transported and stored for laboratory use
                                                          •Provide  for access  to the contained volatiles for  headspace
                                                           analyses
                                                            Commercially-available sampling equipment, such as down-hole
                                                          sampling bombs used  in oil and gas  well drilling, were  reviewed
                                                          but no applicable units were found. Special modifications to com-
                                                          mercially available 2 gal stainless steel pressure vessels were made
                                                          for collecting  contained-volatiles  samples.  Because samples  re-
                                                          mained in the collection vessel until laboratory analyses were com-
                                                          plete, one vessel was required for each  contained-volatiles sample.
                                                            A schematic diagram of this sampling device, termed a remotely-
                                                          actuated  single  point sampler (RASP), is shown in Figure 5. This
                                                          device was equipped with a combination vacuum/pressure gage, a
                                                          pressure  relief  valve pre-set at the vessure pressure rating which
                                                          was much higher than any anticipated pressures, a septum for ob-
                                                          taining headspace vapor samples and  a thermowell. Viton gasket-
                                                          ing was used to seal the removable lid, opened to transfer samples
                                                          to additional containers only after completion  of initial headspace
                                                          analyses and collection of aliquots for additional analyses of raw
                                                          samples.  Manual  and pneumatic valves constructed  of PVDF-
                                                          Kynar were used for isolation and remote operation. The  pneu-
                                                          matic valve  was remotely operated using nitrogen through Teflon
                                                          hoses jacketed in flexible woven stainless steel.
                                                            Prior  to   field   mobilization,  each  completed  sampler was
                                                          thoroughly solvent cleaned, rinsed and purged with nitrogen. Then,
                                                          each  was tested for seal integrity  to  maintain both vacuum and
 36
SCREENING TECHNIQUES

-------
pressure conditions. Evacuation and  purging  with nitrogen were
performed in three cycles, with the final purge of greater than ten
volumes of nitrogen gas. Each unit was then pressurized to about
10 psig with nitrogen for storage and shipment and inscribed with
serial identification markings.
  In the field, samplers were operated as follows:
•A vacuum was pulled on the sampler reducing the pressure to
 about  7 psia  just  prior to sampling;  the actuated valve  was
 attached (closed) and the manual valve opened.
•The sampler was lowered to predetermined depth at the sample
 location.
•The actuated valve was opened for  1 to 2 min to allow sample
 collection through pressure equilibrium.
•The actuated valve was closed, the sampler raised and returned
 to shore.
•The manually operated valve was closed; then the actuated valve
 was removed.
•Finally,  the system was decontaminated  and  the sample  was
 labelled, iced and transported to the laboratory under chain-of-
 custody procedures.
  During sample collection, each sample vessel was at the in situ
pressure at the selected  sampling location and depth.  Using the
vacuum/pressure gage and thermowell, the  temperature, pressure
and vessel weight were recorded, indicating the amount of sample
obtained  and the  sampling conditions. Through this procedure,
sample integrity was maintained including containment of volatiles;
only inert nitrogen  gas and vessel surfaces  had contacted the
sample.
  Under controlled laboratory conditions, each sample vessel was
subsequently heated  in a water bath to a desired temperature while
the pressure was monitored. When the desired temperature equilib-
rium was reached, an  actual headspace  sample was withdrawn
through the septum on each vessel and directly analyzed using gas
chromatography. As a result, numerous volatile compounds were
identified which had not been previously reported, and much more
representative  volatiles  characterization was  performed than
analysis of the bulk liquid matrix alone could provide.
  After headspace analyses, sample vessels  were cooled to allow
aliquots to be collected for additional analyses without significant
volatiles release.
Analytical Equipment
  It has been estimated by the USEPA that 40% to 70% of the
samples submitted to the Contract Laboratory Program yield neg-
ative extractable organics  results  (none  present  above detection
limits).  As a result, increased efforts to screen negative samples
are underway to ease massive seasonal analytical backlog and more
effectively utilize sophisticated analytical capabilities.
  Photo-ionization  and   flame-ionization  detector  equipment
(HNu/OVA) for indications of total organic  vapors have been
commonly used for health and safety as well  as general sample
screening.  Zero  instrument  response indicates  an  absence of
organic vapors. A positive instrument response,  however, is not
conclusive evidence of the  presence of toxic organics. Also, most
organic vapor meters accept samples only in the vapor phase. How-
ever, these  equipment types have proven  useful for screening
samples not requiring volatile organic analyses.
  Some available organic vapor analyzers also are capable of spe-
cific constituent identification and quantitation,  although actual
use has generally been limited.  Increasingly, field OVA screening
in the analytical  mode and "close-support" field-portable or
trailer-housed gas chromatography systems will be available. These
systems should be considered not only in site investigations, but
also during cleanup operations  for fast turnaround analyses upon
which remedial action staff  can base field decisions.  Additional
systems which may prove applicable to specific  sites include infra-
red (IR) and fluorometric analyses, where correlations to types and
levels of contamination can be established.
       PRESSURE
       RELIEF
      THERUOWELL
  2-QAL. PRESSURE
  VESSfL 316SS
                                        ACTUATOR GAS TUBING
                                             ACTUATOR CONTROL
                                             VALVES
NITROGEN
C YLIHDER
                 QUICK-DISCONNECT y    * REMOTE VALVE
                            Figure 3
            Remotely Actuated Single Point Sampler (RASP)
ADDITIONAL ANALYTICAL METHODS

  Analytical methods in addition to GC/MS systems for organics
include atomic adsorption or ICAP for inorganics as well as other
traditional wet chemistry which can be used to gather specific in-
formation and  data for  a more effective assessment of  feasible
technologies.
Incineration Parameters

  Thermal destruction through waste incineration is being increas-
ingly considered as  preferable to land disposal.  Of particular im-
portance in initial assessments of incineration feasibility are heat
value,  ash content,  moisture  content, total halogens and  total
chlorine of the waste. An additional test sometimes utilized is ash
fusion.

Elemental Analyses
  Analyses of total carbon, hydrogen, nitrogen, oxygen, sulfur
and phosphorus are often useful to assist in  identifying  higher
molecular weight organics. These constituent analyses are present
in similar ratios for various classes of organic compounds.

Hazardous Waste Characteristics
  Ignitability, corrosivity and  extraction procedure  (EP)  toxicity
testing according to RCRA criteria can be used for categorizing
wastes and generated residuals and  for  determining  the  hazard
classification. Total cyanide can also be useful regarding reactivity
and hazard potential.

Molecular Weight

  A number of methods, some of which are quite sophisticated,
are available for  number-average or weighted-average molecular
weight analyses. These results can be  very useful in identifying
high molecular weight organic compounds, although the value  of
this test is limited for complex mixtures.
Fractionation

  Samples can  be separated into various fractions through dis-
tillation, solvent extraction, etc. Of particular value in solvent sys-
tems is a modified ASTM distillation test using overhead  con-
densers to separate organic mixtures into fractions according  to
boiling point ranges. The fractions can then be analyzed separate-
ly for desired parameters.
  At one  site, this method showed that a complex organic waste
mixture could be separated so that 95% of the most hazardous
chlorinated solvents could be concentrated in 5-10% of the original
weight. Additional  process design information can be gathered
from this  test if distillation to separate fractions appears feasible
for a given waste media.
                                                                                          SCREENING TECHNIQUES
                                                           37

-------
 BEHAVIORAL PARAMETERS
   In addition to analytical approaches, it is often desirable to de-
 termine how a waste material will behave under certain conditions.
 Using standard testing methods, or method developed for a spe-
 cific case, behavioral parameters can usually be defined.

 Thermal Analyses
   Differential thermal  analysis (DTA) is  used to measure  the
 temperature differential with heating between two sample pans,
 one containing  a ballast of aluminum and/or alumina  and  the
 other  holding the material to  be analyzed. Any thermodynamic
 change such  as melting, evaporation or reaction (oxidation,  de-
 composition, etc.) will cause a temperature  differential  between
 the pans and a subsequent exothermic or endothermic peak re-
 corded at the temperature of occurrence. Typically,  this DTA
 process can  be conducted  from 32 to 930°F  at selected and con-
 trolled temperature rise rates, such as 36'T/min.
   Thermal gravimetric analysis (TGA) generally involves place-
 ment of a small (5-10 mg) sample in a platinum pan which is con-
 tinuously weighed on an electrobalance while  the sample is heated
 to 930°F  in a  nitrogen atmosphere.  Volatiles release,  decom-
 position and other reactions are indicated in  terms of weight loss
 as a function of temperature.
   Results of these tests directly indicate sample behavior at various
 temperatures and can  be compared to behavior  of known chem-
                                                          icals for comparison and identification purposes.  In addition, ad-
                                                          mixtures or testing in other atmospheres (air, oxygen, etc.) can be
                                                          utilized to assess impacts of various conditions on the behavior of
                                                          the material.

                                                          Vapor Pressure Versus Temperature
                                                            If reasonably representative contained-volatiles samples can be
                                                          obtained, a plot of the  vapor pressure versus temperature can be
                                                          readily obtained. These data provide information needed for closed
                                                          system handling, such as required system pressure ratings under
                                                          various conditions. In addition, indications of the nature and con-
                                                          tent of volatile compounds in  the waste material can be obtained.

                                                          Viscosity

                                                            Viscosity is an important parameter in many waste handling and
                                                          processing operations. Examples of the need for viscosity data in-
                                                          clude pumpability and feasibility of injection through feed nozzles
                                                          in liquid injection incinerator systems. Measurements of viscosity
                                                          versus temperature  for  heavy organics provides an indication  of
                                                          whether  heating can be used to improve flow properties  as prac-
                                                          ticed in numerous industries (asphalt, crude oil, etc.). Addition of
                                                          selected solvents or petroleum derivatives (i.e., kerosene) can be
                                                          assessed  for modifying flow properties through  viscosity measure-
                                                          ments. Sometimes even small percentage admixtures can have a
                                                          pronounced effect on handling properties.
38
SCREENING TECHNIQUES

-------
           CHEMICAL  COMPOSITION OF  DRUM  SAMPLES
                        FROM HAZARDOUS WASTE SITES

                                      WILLIAM C.  BLACKMAN, JR.
                                       RICHARD L. GARNAS,  Ph.D.
                                         JOHN E. PRESTON, Ph.D.
                                           CHARLENE M. SWIBAS
                                   U.S. Environmental Protection Agency
                                 National Enforcement Investigations Center
                                                Denver,  Colorado
INTRODUCTION

  Samples taken from drums, tanks, other containers and samples
suspected of having high concentrations of hazardous waste from
221 disposal sites in 41 states and one territory have been prepared
for analysis by the  Regulated Substances  Laboratory  (RSL),
USEPA National  Enforcement Investigations  Center (NEIC),
operated under contract to the USEPA Contract Laboratory Pro-
gram (CLP) by Fred C. Hart Associates, Inc. These samples were
taken from a  wide spectrum  of  hazardous waste sites  by the
USEPA and  state personnel  and contractors  retained  by the
USEPA for the conduct  of  hazardous waste  site evaluations.
After preparation in the RSL, these samples were analyzed in en-
vironmental laboratories of the USEPA regions,  the NEIC and in
eight CLP laboratories.
  The data generated by these analyses are believed to generally
represent the chemical content of drums and other waste containers
on hazardous waste sites throughout the nation and provide policy-
makers, industry and regulatory agency managers, investigators
and analysts with a basis for greater confidence in decisions regard-
ing exposure risks to the public and to personnel engaged in haz-
ardous waste site evaluations.
  The data presented  in this paper should be considered  reliable
to the degree of accuracy and precision required by the contracts
under which they were obtained and achievable through  reason-
able quality control checks  in the collection and compilation there-
of. Appendix A is abstracted from the Registry of Toxic Effects of
Chemical Substances (RTECS), published by the National Insti-
tutes of Occupational Health and Safety (NIOSH). The reader is
cautioned that the Appendix A summary is intended to be a general
overview. The original sources referred to by RTECS should be
consulted for specific characteristics of any of the listed chemicals.
The conclusions reached in this paper are those of the authors and
are not policy statements of the Environmental Protection Agency.

Historical Perspective
  In 1979, as the USEPA began the hazardous waste site cleanup
programs, the Agency had relatively minor experience with and
essentially no procedural criteria for the field investigation of sites,
packaging and shipping of potentially hazardous samples or the
laboratory preparation and analysis of samples suspected  of con-
taining high concentrations of hazardous materials. Criteria were
quickly developed, based  upon the limited experience and pro-
fessional judgment available within the Agency. This early guid-
ance was provided to  USEPA and state technical staffs and con-
tractors through a number of procedures manuals, several of which
were  adapted  from   internal procedural documents   of  the
NEIC1'2'3'4 contractor procedures5 and Department of Transpor-
tation regulations.'
  The early guidance documents  reflected the extreme concerns
held by their authors and proponents for the safety of field inves-
tigators, the transportation  industries, laboratory personnel and
the general public. This concern was based upon good understand-
ing of the potential for spills, releases, exposures, fires and ex-
plosions,  but less certain knowledge of preventive measures and
procedures which could be depended upon to prevent such events
under any and all conceivable field, transportation and laboratory
circumstances.
  At the NEIC, the operational manifestation of these understand-
ings was a consistent effort to "err on the side of safety". This pol-
icy permeated site investigation plans; training of NEIC employees
and training provided by NEIC to other USEPA, state and contrac-
tor personnel; and continuously updated sample packaging and
shipment procedures and RSL procedures. Other USEPA elements
adapted or adopted  portions  of  these procedures, modified or
added to  them to meet local and Regional requirements or devel-
oped criteria and procedures  independently. Regardless  of  the
degree or extent of independence,  the conservative philosophy to-
ward safety aspects appears  to  have been generally adopted
throughout the Agency.

ANALYTICAL PROCEDURES

  The samples which are the subject of this paper  were taken from
drummed material, waste pits or ponds, piles of waste, tank trucks
or on-site tanks and contaminated  soils. Many of the samples were
used oil, spent solvents, paint wastes, metal treatment and plating
wastes and polymer formulations. They were usually industrial pro-
cess wastes, waste raw materials and byproducts, synthesis inter-
mediates and off-specification products. The RSL, operating in a
containment laboratory configuration and under strict contain-
ment laboratory  procedures,  received  8 oz hazardous  waste
samples as shipped and provided initially screened and diluted ali-
quots of each defined phase. Many of the samples contained two or
more phases when received by the RSL. Phases were separated and
preparations were made from each  defined phase.
  The organic analytical regime to which the preparations have
been subjected by CLP laboratories has varied somewhat according
to the contract language in force at any given time. From  1980 to
1982, the specified organic analyses included 113 priority pollutants
and a maximum of 30 mass spectrometry library identifications
(tentative identifications). From 1982 to the present, the organic
analyses included an additional 20 non-priority pollutant organic
substances. The categories included 11 priority pollutant organic
                                                                                    SCREENING TECHNIQUES
                                                       39

-------
acids, 45 priority pollutant organic bases and neutrals, 31 priority
pollutant organic volatile solvents, 26  priority pollutant organic
pesticides and polychlorinated biphenyls (PCBs) and the 20 addi-
tional organics mentioned above (Table 1).
  The contract  requirement for analysis of these organics, which
are detectable by gas chromatography, imposes certain limitations
which should be recognized by the reader. Many common indus-
trial chemicals either have poor extraction efficiencies or do not
chromatograph. Polymers, carbor.ylates, glycols, sulfonates, phos-
phates and low molecular weight  alcohols, amines and  aldehydes
will not be qualitatively or quantitatively analyzed by the methods
specified.  In addition, the pesticides/PCB analysis includes only
priority pollutant chlorinated hydrocarbon pesticides and excludes
carbamate and organophosphate  insecticides and  nitrogenous  or
phenoxy herbicides. Constituents of these categories may  have been
identified  by analytical procedures available under the later Special
Analytical Services (SAS) contracts, when specified, but the great
majority of samples reported upon in this paper were not subjected
to these additional analyses.
   The inorganic analyses were performed by inductively coupled
argon plasma spectroscopy (ICP), atomic adsorption spectroscopy
(AA) and colorimetry. Approximately 300 samples were subjected
to procedures which identified and quantified 13 priority pollutant
elements,  cyanide and 20 additional inorganic  elements. Approxi-
mately 1,300 samples were analyzed for 35 inorganics.

ANALYTICAL RESULTS

   The  analytical procedures described  yielded  the organic data
summarized in  Table 2 and the inorganic data similarly provided
in Table 3. The average number  of analyses for organic constit-
uents was 1,100. These constituents were detected  by only  3%  of
the analyses (detected/analyzed). The average number of analyses
for inorganic constituents was similarly 1,200 but,  in contrast, the
inorganics were detected by 39% of the analyses.
   The  average of the mean concentrations of the 114 reported
values was 576  mg/1, while the average mean concentration of the
36 reported values greater  than  100 mg/1 was 1,728 mg/1. The
organic chemicals with the highest reported maximum concentra-
tions (in percentages) are 2-methylphonol (90%);  trichloroethene
(82%); o-xylene (79%); chlordane (78%); acetone (76%);  1,1,1-
trichloroethane (72%); and  benzene (60%).  Significantly, 39%  of
                                                            the 133 organic chemicals were detected in only four or fewer sam-
                                                            ples.  Moreover, 64 organic constituents were detected in less than
                                                            one of every 100 samples.
                                                               The average of the mean concentrations  of the 35 inorganic
                                                            constituents, all of which were detected in some samples, was 11,836
                                                            mg/1, while the average mean concentration for the 16 reported
                                                            values greater than 100 mg/1 was 3,876 mg/1. The highest reported
                                                            maximum  inorganic  concentrations  were iron (94%);  sodium
                                                            (86%); zinc (75%); lead (66%); silicon (38%); and calcium (35%).
                                                            The inorganics detected in the fewest samples were cyanide (2.7%
                                                            of times analyzed) and thallium (8.0% of times analyzed).
                                                               Finally, more than 450 additional  non-target organic constit-
                                                            uents were tentatively identified or quantified. A listing of these
                                                            constituents may  be obtained by contacting one of the authors at
                                                            the NEIC.

                                                            SIGNIFICANCE OF CHEMICAL
                                                            CONSTITUENTS IDENTIFIED
                                                               An exhaustive  evaluation of the significance of each detected
                                                            chemical constituent is not possible in this paper. However, a sum-
                                                            marized tabulation which will enable the reader to quickly gain a
                                                            sense of the general nature of any of the 133  organic and 35 inor-
                                                            ganic target constituents is given in Appendix A. The Appendix A
                                                            indicators are: (1) priority pollutant per the NRDC v. Train con-
                                                            sent  decree;' (2) includion in the Department of Transportation
                                                            (DOT) regulations pertaining to transport of hazardous materials;*
                                                            (3) chemical or compound for which an Occupational Safety and
                                                            Health Administration (OSHA) standard pertains;' (4) one or more
                                                            of the RCRA/CERCLA indicators'—EP toxicity, ignitability, per-
                                                            sistence, reactivity, corrosiveness;  (5) severe toxicity to a test ani-
                                                            mal;1 (6) severe reproductive effect;* (7) severe irritation (skin or
                                                            eye);' (8)  known carcinogen;' (9) mutagen;' or (10)  teratogen.1
                                                            Some general  conclusions regarding environmental hazards, ex-
                                                            posure risk to  field investigators, exposure risk to laboratory per-
                                                            sonnel and shipment  of hazardous  waste samples, are offered in
                                                            subsequent sections of this paper.

                                                            Environmental Significance
                                                               A  sense of the chemicals present in  more plentiful quantities on
                                                            hazardous waste  sites may be  had by the weighting scheme pro-
                                                            vided in Tables 4 (organics) and 5 (inorganics). Mean concentra-
                                                               Table 1
                                                Target Chemical Constituents by Category
  ORGANIC ACIOS

   I  2,4,6-Trlchloroohenol   5
   2  p-Chlaro-eHCreeol      6
   3  2-Chlorophenol        7
   It  2,4-Dlchlorophenol      I

  ORGANIC BASES AMD HEL/TBALS
                                                                      PESTICIDES. K3» AND TOO
                 2,4-Dl«*chylphenol
                 2-NUrophen>L
                 4-NUrophenol
                 2.4-Olnlcrophenol
12
U
14
15
16
17
1ft
19
20
21
22
23
24
25
26
Acenaphthene
Benzldlne
1 . 2 ,4-Trlchlorobemene
Hexachlorobenzene
Hexachloroechane
Bli(2-Chloroechyl)Łth*r
2-Chloronaphthalene
1 ,2-Dlchlorobenzane
1.3-Dlchlorobenzene
1.4-Dlchlorobenzene
3.3'-Olchlorobenzldlne
2, 4-Dlnltro toluene
2,6-Dlnltrotoluene
1 , 2-Dlpnenylhydrazlne
Fluoranthene
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
4-chlorophenyl Phenyl Ether
4-Broraphenyl Phenyl Ether
Ble(2-ChloroUopropyl)Ether
Bll(2-Chloro«thoxy;Hethane
Hexachlorobutad lene
HeMachlovocyclopcncadtene
Iiophorone
Naphthalene
Nitrobenzene
N-Nltroaodlphenylanlne
N-NUroaodlpropylolne
Bl>(2-Ethylhexyl)rnthalate
Benzyl Butyl Pnchalate
Dl-n-8utyl PhthaUte
Dl-n-Occyl Phthalate
  ORGANIC VOLATlLeS

  57 Acroleln            68
  58 Acrylonlcrllc        69
  59 Benzene             70
  60 Carbon Tecrichlorlde    71
    Chlorobemene        72
    l,2-Dlchloro
-------
                                                             Table 2
                                    Results of Hazardous Waste Analyses Organic Chemical Constituents
CAS
88-06-2
59-50-7
95-57-8
120-83-2
105-67-9
88-75-5
100-02-7
51-28-5
534-52-1
87-86-5
108-95-2
83-32-9
92-87-5
120-82-1
118-74-1
67-72-1
111-44-4
91-58-7
95-50-1
541-73-1

106-46-7
91-94-1
121-14-2
606-20-2
122-66-7
206-44-0
7005-72-3
101-55-3
39638-32-9
111-91-1

87-66-3
77-47-4
78-59-1
91-20-3
98-95-3
86-30-6
621-64-7
117-81-7
85-68-7
84-74-2
117-84-0
84-66-2
131-11-3
56-55-3
50-32-8
205-99-2
207-08-9
218-01-9
208-96-8
120-12-7
191-24-2
86-73-7
85-01-8
53-70-3
193-39-5
129-00-0
107-02-8
107-13-1
71-43-2
56-23-5
108-90-7
107-06-2
71-55-6
75-34-3
79-00-5
79-34-5
75-00-3
NUMBER OF TIMES CONCENTRATION (PPM)
COMPOUND NAME ANALYZED DETECTED MEAN MAXIMUM
2,4,6-TRICHLOROPHENOL
p-CHLORO-m-CRESOL
2-CHLOROPHENOL
2.4-OICHLOROPHENOL
2.4-DIMETHYLPHENOL
2-NITROPHENOL
4-NITROPHENOL
2,4-DINITROPHENOL
4,6-DINITRO-o-CRESOL
PENTACHLOROPHENOL
PHENOL
ACENAPHTHENE
BENZIDINE
1,2,4-TRICHLOROBENZENE
HEXACHLOROBENZENE
HEXACKLOROETHAUE
BI S( 2-CHLOROETHYL 1ETHER
2-CHLORONAPHTHALENE
1,2-DICHLOROBENZENE
1.3-DICHLOROBENZENE

1,4-DICHLOROBENZENE
3.3'-DICHLOROBENZIOINE
2,4-DINITROTOLUENE
2,6-DINITROTOLUENE
1,2-DIPHENYLHYDRAZINE
FLUO RAN THEN E
4-CHLOROPHENYL PHENYLETHER
4-BROMOPHENYL PHENYL ETHER
8IS(2-CHLOROISOPROPYL)ETHER
BIS(2-CHLOROETHOXY)METHANE

HEXACHLOROBUTADIENE
HEXACHLOROCYCLOPENTADIENE
ISOPHORONE
NAPHTHALENE
NITROBENZENE
N-NITROSODIPHENYLAMINE
N-NITROSODIPROPYLAMINE
BIS(2-ETHYLHEXYL)PHTHALATE
BENZYL BUTYL PHTHALATE
DI-n-BUTYL PHTHALATE
DI-n-OCTYL PHTHALATE
DIETHYL PHTHALATE
DIMETHYL PHTHALATE
BENZO(a)ANTHRACENE
BENZOIalPYRENE
BENZOIblFLUORANTHENE
BENZOUIFLUORANTHENE
CHRYSENE
ACENAPHTHYLENE
ANTHRACENE
BENZOtghDPERYLENE
FLUORENE
PHENANTHRENE
DIBENZO(a,h)ANTHRACENE
INDEKO(l,2,3-cd)PYRENE
PYRENE
ACROLEIN
ACRYLONITRILE
BENZENE
CARBON TETRACHLORIOE
CHLOROBENZENE
1,2-CICHLOROETHANE
1,1.1-TRICHLOROETHANE
1,1-DICHLOROETHANE
1,1,2-TRICHLOROETHANE
1 ,1 ,2 ,2-TETRACHLOROETHANE
CHLOROETHANE
1131
1131
1131
1131
1131
1131
1131
1131
1131
1137
1131
1256
1132
1250
1273
1250
1150
1250
1250
1250

1250
1239
1250
1250
1136
1256
1239
1250
1239
1250

1250
1272
1250
1256
1250
1255
1250
1250
1250
1250
1250
1250
1250
1255
1248
1228
1243
1256
1252
1256
1250
1256
1256
1250
1239
1255
1052
1052
1225
1225
1225
1225
1225
1225
1217
1225
1225
0
3
3
3
50
1
0
0
0
16
127
30
I
16
30
3
4
2
50
14

25
0
2
1
0
65
0
0
2
1

4
28
27
168
2
16
1
190
49
97
26
47
31
50
36
29
23
46
22
77
14
58
110
3
14
79
0
2
104
11
50
33
98
15
10
10
1
0.41
0.73
0.08
637.68
0.07
666.97
2406.95
26.96
0.00
66.72
123.10
8.57
1.97
0.30
1936.82
387.84

993.04

0.06
0.06
_
66.87

_
0.09
0.06

1.82
58.72
158.23
269.02
0.24
95.08
0.04
213.56
88.55
509.09
1.69
25.29
51.37
28.46
10.87
22.97
22.02
27.72
4.28
218.47
1.45
45.56
227.92
0.09
2.24
50.44
19.96
577.29
341.99
85.38
416.69
1318.02
12.50
307.25
256.09
0.05
310.00
820.00
65.00
140000.00
76.00
370000.00
400000.00
8400.00
2.20
36000.00
21000.00
8300.00
1800.00
350.00
490000.00
220000.00

540000.00

78.00
78.00

42000.00

_
'87.00
78.00

2000.00
14000.00
160000 .00
81000.00
220.00
110000.00
50.00
45500.00
50000.00
560000.00
520.00
4500.00
30000 .00
21600.00
7200.00
10200 .00
10200.00
21600.00
3800 .00
126000.00
400.00
16800.00
126000.00
98.00
630.00
33000.00
21000.00
600000 .00
400000.00
57000.00
270000 .00
720000.00
5000.00
240000.00
310000.00
57 .00
CAS
110-75-8
67-66-3
75-35-4
156-60-5
78-87-5
10061-02-6
10061-01-5
100-41-4
75-09-2
74-87-3
74-83-9
75-25-2
75-27-4
75-69-4
75-71-8
124-48-1
127-18-4
108-88-3
79-01-6
75-01-4
309-00-2
60-57-1
57-74-9

50-29-3
72-55-9
72-54-8
115-29-7
115-29-7
1031-07-8
72-20-8
7421-93-4
76-44-8
1024-57-3

319-84-6
319-85-7
319-86-8
58-89-9
53469-21-9
11097-69-1
11104-28-2
11141-16-5
12672-29-6
11096-82-5
12674-11-2
8001-35-2
1746-01-6
65-85-0
95-48-7
108-39-4
95-95-4
62-53-3
100-51-6
106-47-8
132-64-9
91-57-6
88-74-4
99-09-2
100-01-6
67-64-1
78-93-3
75-15-0
591-78-6
108-10-1
100-42-5
108-05-4
95-47-6
COMPOUND NAME
2-CHLOROETHYLVINYL ETHER
CHLOROFORM
1,1-DICHLOROETHENE
TRANS-1 ,2-DICHLOROETHENE
1,2-DICHLOROPROPANE
TRANS-1 ,3-DICHLOROPROPENE
CIS-1,3-DICHLOROPROPENE
ETHYLBENZENE
METHYLENE CHLORIDE
CHLOROME THANE
BROMOME THANE
BROMOFORM
BROMOO ICHLOROMETHANE
FLUOROTRICHLOROMETHANE
DICHLOROOIFLUOROMETHANE
CHLOROOIBROMOMETHANE
TETRACHLOROETHENE
TOLUENE
TRICHLOROETHENE
VINYL CHLORIDE
ALDR1N
OIELORIN
CHLORDANE

4. 4 '-DOT
4, 4' DDE
4 ,4 'ODD
a-ENOOSULFAN
b-ENDOSULFAN
ENDOSULFAN SULFATE
ENORIN
ENORIN ALDEHYDE
HEPTACHLOR
HEPTACHLOR EPOXIOE

a-BHC
b-BHC
d-BHC
g-BHC (LINDANE)
PCB 1242
PCB 1254
PCB 1221
PCB 1232
PCS 1248
PCB 1260
PCB 1016
TOXAPHENE
2,3,7.8-TETRACHLORO-
DIBENZO-p-DIOXIN
BENZOIC ACID
2-METHYLPHENOL
4-METHYLPHENOL
2,4,5-TRICHLOROPHENOL
ANILINE
BENZYL ALCOHOL
4-CHLOROANILINE
DIBENZOFURAN
2-METHYLNAPHTHALENE
2-NITROANILINE
3-NITROANILINE
4-NITROANILINE
ACETONE
2-BUTANONE
CARBONDISULFIDE
2-HEXANONE
4-METHYL-2-PENTANONE
STYRENE
VINYL ACETATE
o-XYLENE
NUMBER OF TIMES CONCENTRATION (PPM
ANALYZED DETECTED MEAN MAXIMUM
1141
1225
1225
1225
1225
1140
1140
1225
1225
1224
1225
1225
1225
1184
1064
1224
1225
1225
1225
1224
1143
1142
1142

1142
1142
1142
1142
1141
1141
1142
L101
1142
1142

1142
1142
1142
1142
1142
1145
1142
1142
1145
1144
1143
1142
972
508
498
498
498
499
499
499
502
503
499
499
499
510
508
510
510
512
510
510
600
0
29
22
18
22
3
0
375
208
4
1
0
0
5
1
0
150
469
106
4
11
18
41

16
20
14
13
2
2
6
2
48
10

10
9
4
9
27
34
1
1
18
37
6
2
6
16
44
28
0
0
0
2
9
35
0
0
0
40
54
2
14
46
26
6
223
20.49
14.11
3.86
91.50
0.02
2279.25
780.83
0.38
0.01
80.15
0.94
-
1345.26
10208.85
2173.11
0.15
2.00
44.73
2548.44

0.21
0.04
0.11
0.04
0.06
0.08
0.01
0.08
115.90
0.01

0.69
0.07
0.04
0.35
2.80
5.21
0.00
0.00
8.87
390.36
0.12
1.05
0.00
131.64
3827.72
1128.10
0.00
38.80
89.35
6651.64
6902.96
0.23
1945.97
1170.95
1742.23
231.18
8388.21
14000.00
9300.00
4300.00
40000.00
10.00
150000.00
220000.00
337.00
10.00
45000.00
1000.00
-
170000.00
440000.00
820000.00
180.00
2000 .00
33000 .00
780000.00

100.00
23.00
63.00
22.00
66.00
86.00
2.80
84.00
110000.00
6.80

400.00
35.00
21.00
260.00
1600.00
3620.00
0.01
0.01
8000 .00
140000.00
119.00
1200.00
0.60
20000.00
900000.00
110000.00
1.00
12000.00
10000.00
760000.00
565000.00
100.00
490000.00
300000.00
300000 .00
58000 .00
790000.00
tion values have been converted to percent. Frequency detected is
the number of times detected divided  by the number of times
analyzed.  The product  of the mean concentration and the fre-
quency detected (X x F) yields a weighted frequency which  rep-
resents the equal importance of these two variables. To facilitate
ranking, the greatest weighted frequency (toluene and silicon) have
been normalized to 100, and the other values have been adjusted
accordingly.
  The most prevalent 20 organic constituents have been ranked in
Table 4. Based upon weighted frequency, the five most prevalent
organics were toluene > o-xylene  > 2-butanone > ethylbenzene
> acetone. As indicated in Appendix A, two are priority pollu-
tants, four are DOT regulated, four have applicable OSHA stan-
dards, all are toxics, four are ignitable, three are persistent, one is
a mutagen and none are known carcinogens. These  five organic
constituents represent about 81 %, by total weighted frequencies, of
the total 133  target organics. The weighted frequencies of the re-
maining 15 organic constituents ranged from less than 1 % to 9% of
the weighted frequency for toluene.
  The 20 prevalent organics of Table 4 included 12 priority pollu-
tants, 17 DOT regulated substances, 18  OSHA regulated constit-
uents, 18 toxics, 13 ignitables, 15 persistents, 14  constituents ex-
hibiting severe reproductive effects, 13 severe irritants, 11 carcin-
ogens, ten mutagens and seven teratogens. In fact, 12 of the 20 ex-
                                                                                          SCREENING TECHNIQUES
                                                          41

-------
hibited nearly all of the Appendix A characteristics. These 20
organic constituents represent in excess of 97%, by total weighted
frequencies, of all 133 target organic constituents.
In similar fashion, the 15 most prevalent inorganic constituents
have been ranked in Table 5. Based upon weighted frequency,
the seven most prevalent inorganic elements were silicon > iron
^ calcium ^ sodium > aluminum > potassium ?>• titanium.
These seven more prevalent inorganics include no priority pollu-
tants nor known carcinogens. The seven represent nearly 90%, by
total weighted frequencies, of all 35 target inorganics.
The weighted frequencies of the remaining eight elements ranged
from less than 1% to 9% of the weighted frequency for silicon.
This group of eight, having less than 10% weighted frequencies, in-
cludes the five inorganic priority pollutants found on Table 5: zinc,
lead, chromium, copper and cyanide.
Table 3
Results of Hazardous Waste Analyses Inorganic Chemical Constituents
NUMBER OF TIMES CONCENTRATIONS (PPM)
COMPOUND NAME ANALYZED DETECTED MEAN MAXIMUM
ALUMINUM(AL) 1311 737 4621.88 252000.00
ANTIMOUY(SB) 1411 470 80.53 14400.00
ARSENIC(AS) 1492 507 114.45 143850.00
BARIUM(BA) 1363 620 476.98 143000.00
BERYLLIUM(BE) 1532 216 LIB 466.00
BORON(B) 1023 272 62.27 23400.00
CADMIUM(CD) 1565 444 14.02 5100.00
CALCIUM(CA) 1154 731 5705.22 353000.00
CHROMIUM(CR) 1574 834 806.30 312000.00
COBALT(CO) 1275 334 11.95 1110.00
COPPER(CU) 1529 877 521.81 95400.00
CYANIDE(CN) 1200 33 298.00 105000.00
IRON(FE) 1315 994 11674.65 938000.00
LANTHANUM! LA) 832 95 4.47 1150.00
LEAU(PB) 1578 780 2130.92 656000.00
MAGNESIUM(MG) 1152 697 1153.15 134000.00
MANGANESE(MN) 1320 842 155.31 43500.00
MERCURY(HG) 1167 291 3.17 1900.00
MOLYBOENUM(MO) 1086 148 91.01 17300.00
NICKEL(NI) 1533 565 146.85 156000.00
POTASSIUM(K) 446 297 3110.48 87500.00
SCANDIUM) SO 963 167 0.53 18.50
SELENIUM(SE) 1417 261 11.83 1900.00
SILICON! SI) 967 566 20411.47 378000.00
SILVER(AG) 1553 207 1.52 271.00
SODIUM(NA) 1152 549 6966.18 856000.00
STRONTIUM(SR) 1081 602 43.54 6390.00
THALLIUM! TL) 1367 110 1.76 400.00
TIN(SN) 422 54 18.78 4100.00
TITANIUM(TI) 1084 610 <:502.58 244000.00
TUNGSTEN(U) 942 127 19.94 13100.00
VANADIUH(V) 1315 358 6.92 365.00
YITRIUM(Y) 1046 195 4.09 372.00
ZINC(ZN) 1532 1105 1520.36 745000.00
ZIRCONIUM(ZR) 950 261 18.00 2190.00
The 15 Table 5 inorganics include the five priority pollutants,
eight elements subject to DOT regulations, three elements for
which OSHA standards exist, three RCRA/CERCLA toxics and,
depending upon the specific form or compound present, several
ignitables, reactives and corrosives. Also present are three inor-
ganics having reproductive effects, two known carcinogens, two
mutagens and two teratogens. These 15 inorganic constituents rep-
resent in excess of 99%, by total weighted frequencies, of all 35
target inorganic constituents.
Clearly, the chemical constituents present in drums, other con-
tainers and in contaminated soil on hazardous waste sites consti-
tute potential or actual hazards to the environment and to the pub-
lic health. The principle concern in the preparation of the early
guidance1 • 2i 3 was the potential for contamination of aquifers which
supply domestic water systems. This review lends no rationale for
Table 4
Prevalent Organic Constituents
(X)(F)
Mean Frequency (tomallMd
Concentration Detected to
Constituent X (JQ X (F) (X)(F) Toluene
Toluene 1.021 38.3 39.LO 100
o-Xylene 0.839 37.2 31.20 80
2-Butanone 0.690 10.6 7.31 19
Ethylbenzene 0.228 30.6 6.98 18
Acetone 0.665 7.8 5.21 13
2-Methylphenol 0.383 8.8 3.39 9
Phenol 0.241 11.2 2.70 7
Trlchloroethene 0 217 87 1.88 5
Tetrachloroethene O.L35 12.2 1.65 4
Mechylene Chloride 0.078 17.0 1.33 3
1.1.1-Trlchloroethane 0.132 8.0 1.06 3
4-Methyl-2-Pentanone 0.117 9.0 1.05 3
Chlordane 0.255 3.f 0.92 2
Styrene 0.174 5.1 0.89 2.
1,2-Dlchlorobenzene 0.194 4.0 0.78 2
4-Methylphenol 0.113 5.6 0.64 2
2-Hexanone 0.195 2.8 0.54 1
Benzene 0.058 8.5 0.49 1
Naphthalene 0.027 13.4 0.36 <1
Bls(2-Ethylhexyl)Phthalate 0.021 15.2 0.32 <1
diminished concern in that regard. A wide variety of the prevalent
organic constituents are persistent in the environment, are frequent
groundwater pollutants and are shown to be present in such con-
centrations that, even with extremes of dilution, unacceptable con-
centrations could be expected to remain.
Some constituents, such as cadmium, mercury, endrin and lin-
dane, have maximum contaminant levels (MCL) in drinking water
in the low ug/1 range.10 Carbon tetrachloride, tetrachlorethene, tri-
chloroethene, vinyl chloride and benzene have recommended max-
imum contaminant levels (RMCL) set at zero." Aldrin, dieldrin,
toxaphene and benzo(a)pyrene have ambient water quality criteria
(AWQC) in the low ng/1 range." These compounds were detected
a total of 1,192 times. The inevitable deterioration of drums and
other waste containers constitutes a potential long term hazard to
the groundwater resources of the nation.
Recent literature indicates that Volatile Organic Chemicals
(VOCs) such as tetrachloroethene, trichloroethene and dichloro-
ethene are biotransformed in groundwater to refractory com-
pounds such as vinyl chloride. "• l4 Vinyl chloride exhibits nearly
all of the Appendix A characteristics and is a particularly potent
carcinogen. Vinyl chloride is being detected in groundwater which
has been contaminated by the precursor VOCs noted above."
These VOCs and others were detected in more than 10% of the
analyses, and one sample contained 82% trichloroethene.
The inorganic constituents identified are conspicuous because of
the high frequencies of detection. Many of these elements may now
be considered commonplace on hazardous waste sites. Again, de-
pending upon the form in which deposited or leached, nearly all
are threats to groundwater supplies. Metals such as mercury, lead,
cadmium and chromium, in soluble compounds at very low con-
centrations, are potent long-term health hazards in groundwater.
42
SCREENING TECHNIQUES

-------
  The findings confirm the magnitude of the threat to groundwater
supplies of deteriorating drums and tanks, leaking pits and ponds
and leaking land disposal sites containing these chemicals. More-
over, they strengthen the likelihood that unknown, abandoned, im-
properly sealed  and improperly closed  disposal sites can be ex-
pected to continue threatening  groundwater  supplies for many
years.

Significance for Field Investigators
  While all  of the characteristics of the priority pollutants are of
general concern  to the field investigator, those which are immed-
iate, on-site hazards are of specific intense concern. These include
immediate toxicity, ignitability,  reactivity,  corrosivity and severe
irritation. Several of the chemicals found are extremely toxic, e.g.,
dinitrophenol, aldrin  and TCDD (tentatively identified). Among
the five most prevalent organics, four are ignitable. Maximum con-
centrations  among  the  prevalent five ranged from  15  to 79%.
Among the  entire 133 organics, RCRA/CERCLA toxics were de-
tected 3,997 times,  ignitables were detected 7,739 times, reactives
were present in 371  samples and severe eye and skin or respiratory
irritants were present in 1,609 samples. 2, 3, 7, 8-TCDD was tenta-
tively identified at low concentrations in 6 samples. Inorganic con-
stituents, depending upon the species in which present, also consti-
tute toxic,  ignitability  (e.g., elemental  sodium),  reactivity  and
severe irritant hazards. Cyanide  was present in 33 samples. Thus,
encounter with a material having one or more of these properties,
is likely at any time a container on a hazardous waste site is opened.
Moreover, since the CLP contracts do not require analyses for
acutely toxic organophosphates or carbamates, their presence can-
not be ruled out. There is no basis in these data, for any relaxation
of the onsite safety procedures referenced earlier, particularly those
dealing with opening and sampling of containers.

Significance in Packaging and Shipping of Samples
  Considerations pertinent  to  packaging and shipment of  sus-
pected hazardous waste samples are essentially identical to those of
the field investigator.  These findings reinforce the 1979 and subse-
quent EPA  Office of Health and Safety (OHS) guidance directing
adherence to 49  CFR  172-173 requirements. The data further indi-
cate the imperative that packaging be accomplished to preclude any
possibility of leakage, breakage, or contact by incompatibles. OHS
will shortly issue newly clarified  "National Guidance for Comply-
ing with DOT Regulations in Shipping Hazardous Samples".

 Significance for Laboratory Personnel
   Managers,  supervisors, analysts, and support personnel staffing
 laboratories that perform analyses  on hazardous waste samples
 share the field investigator's concerns with the immediate hazards
 of toxic, ignitable, reactive, and irritant materials. If careless or in-
 adequately  trained,  clothed or equipped, laboratory  personnel
 may incur long-term  risk from exposure to materials that are car-
 cinogens  or mutagens or  that exhibit teratogenic, reproductive  or
 severe irritant properties. A consideration of major operational
 significance is the prevention  of contamination of  work areas,
 equipment and instruments in environmental level laboratories.
   The field investigator may find it necessary to deal with large
 quantities of waste while conducting sampling and other field oper-
 ations. In contrast, since Agency practice has standardized around
 shipment of 8 oz sample containers, the quantity of material  to
 which laboratory personnel may come into contact is significantly
 reduced.  Leaks, spills,  or ignition of such quantities of the waste
 materials identified herein, may be dealt with safely in a properly
 designed and operated hood. The findings herein support the con-
 cept  that laboratory personnel, if  properly trained, supervised,
 equipped and clothed, may perform operations incident to prepara-
 tion of 8 oz waste samples for analysis without incurring risks be-
 yond those  assumed in normal environmental level laboratory oper-
 ations. These findings  do not provide a basis for relaxing safety
 rules or guidance or for  attempting short cuts in laboratory pro-
 tocols or procedures.
                            Table 5
                 Prevalent Inorganic Constituents
Mean
Concentrat Ion
Constituent % (X)
Silicon
Iron
Calciun
Sod inn
Aluninun
Potass ion
Titaniun
Zinc
Lead
Magnesium
Chroraiun
Copper
Bar tun
Manganese
Cyanide
2.041
1.167
0.571
0.697
0.462
0.311
0.250
0.152
0.213
0.115
0.081
0.052
0.048
0.016
0.030
Frequency
Detected
% (F)
58.5
75.6
63.6
47.7
56.2
66.6
56.3
72.1
49.4
60.5
53.0
57.4
45.5
63.8
2.8
(X)(F)
Normalized to
^X) (F) Silicon
119.5
88.2
36.2
33.2
26.0
20.7
14.1
11.0
10.5
7.0
4.3
3.0
2.2
1.0
0.08
100
74
30
28
22
17
12
9
9
6
4
3
2
1
<1
  Many of the wastes (e.g., VOCs and heavy metals) present in
environmental samples in high concentration can contaminate large
areas  of the laboratory  and its personnel and equipment. Also,
some  of these  pollutants are extremely toxic or present long-term
hazards such as cancer or mutagenic change. For these reasons, the
practice of using a modern,  well equipped, separate laboratory
for sample preparation should be continued.

CONCLUSIONS
  Data from the organic and inorganic analyses of samples from
drums and other containers, on hazardous waste sites throughout
the nation, show the presence of priority  pollutants, DOT and
OSHA regulated substances, constituents having severe reproduc-
tive effects,  severe irritants, carcinogens,  mutagens, and terato-
gens,  in significant numbers of identifications and in concentra-
tions  as high as 93%. These data indicate the magnitude of the
threat to groundwater supplies of deteriorating drums and tanks,
leaking pits and ponds and leaking land disposal sites containing
these  chemicals. They strengthen the likelihood that unknown,
abandoned, improperly sealed and improperly closed disposal sites
can be expected to continue threatening groundwater supplies for
years to come.
  The early procedural guidance, adopted by EPA, for the  field
investigation of hazardous waste sites and packaging of samples for
shipment, is shown to be valid. The data confirm that laboratory
personnel, working in a properly designed containment facility, if
properly trained, supervised, equipped and  clothed, may perform
operations incident to preparation of 8 oz waste samples for analy-
sis without incurring risks  beyond those assumed  in normal en-
vironmental level laboratory operations.  A well designed, main-
tained and operated separate laboratory will provide the necessary
protection from contamination  of nearby or adjacent environ-
mental level laboratories.

ACKNOWLEDGEMENTS

  The authors gratefully acknowledge the assistance and/or con-
structive criticisms of Earl W. Beam, Marcia L. Colvin, Thomas P.
                                                                                          SCREENING TECHNIQUES       43

-------
 Gallagher, Donald C. Gipe, James L. Hatheway, Robert H. Laid-
 law, Dennis E. Longsine,  Dr. Joe H. Lowry, Dr. Theodore O.
 Meiggs, Floy E.  Park,  Mary F. Rohrer,  Robert  F. Schneider,
 Richard W. Warner and Gary D. Young.

 REFERENCES

   1. U.S. Environmental  Protection  Agency, Office  of  Enforcement,
     National Enforcement Investigations Center, Procedures for the Field
     Investigation of Uncontrolled Hazardous Waste Sites.  Denver,  CO,
     1979.
   2. U.S. Environmental Protection Agency, Office of Health and Safety,
     Safety Manual for Hazardous Waste Site Investigations, Scpl. 1979.
   3. U.S. Environmental Protection Agency, Office of Enforcement, Na-
     tional Enforcement Investigations Center, Enforcement  Considera-
     tions for Evaluations  of Uncontrolled Hazardous  Waste Disposal
     Sites by Contractors, Denver, CO, Apr.  1980.
   4. U.S. Environmental  Protection  Agency, Office  of  Enforcement,
     National  Enforcement Investigations  Center, Regulated  Substances
     Laboratory Manual, Denver, CO, Apr. 1980.
   5. Procedures Manual for NE1C Regulated Laboratory Operations. Den-
     ver, CO, Fred C. Hart Associates, Inc., May 1982.
   6. 49 CFR 172-173.
                                                             7. Consent Decree in NRDCv. Train.
                                                             8. U.S. Department of Health and Human Services, National Institutes
                                                               of Occupational Safety  and  Health,  Registry of Toxic Effects of
                                                               Chemical Substances. 1-3.
                                                             9. 40 CFR 281, 40 CFR 300.
                                                            10. 40 CFR 141.
                                                            11. 49/^24330, June 12, 1984.
                                                            12. 49/7*79318. Nov. 28, 1980.
                                                            13. Frances Parsons, Paul R. Wood and  Jack DcMarco, "Transforma-
                                                               tions of Tetrachloroethene and Trichloroethene in  Microcosms and
                                                               Groundwater," J.  of Ike Amer.  Water Works Association,  1984,
                                                               56-59.
                                                            14. Jack DeMarco, "History of Treatment of Volatile Organic Chemicals
                                                               in Water," Occurrence and Removal of Volatile Organic Chemicals
                                                               from Drinking  Water. Denver, CO, AWWA  Research Foundation,
                                                               1983, 1-29.
                                                            15. U.S. Environmental Protection Agency, Office of Enforcement and
                                                               Compliance Monitoring, National Enforcement Investigations Center,
                                                               South Florida Drinking Water Investigation Broward, Dade and Palm
                                                               Beach  Counties,  by James R. Vincent, Report No. EPA-330/1-84-
                                                               001, Denver, CO, June 1984.
44
SCREENING TECHNIQUES

-------
                        SURVEY OF MOBILE  LABORATORY
                    CAPABILITIES AND  CONFIGURATIONS

                                                   J.L. ENGELS
                                                 H.B. KERFOOT
                                                  D.F. ARNOLD
                                              R.H. PLUMB, Ph.D.
                           Lockheed Engineering and Management Services, Inc.
                                                Las Vegas,  Nevada
                                               S.  BILLETS, Ph.D.
                                    U.S. Environmental Protection Agency
                               Environmental Monitoring Systems Laboratory
                                                Las Vegas,  Nevada
 INTRODUCTION

   An essential element in the management of uncontrolled haz-
 ardous waste sites is analytical support service to determine the
 hazardous substances  present, the extent of environmental con-
 tamination that has occurred and the effectiveness of any cleanup
 efforts that have  been undertaken. Because of the variability in
 site location, the time lag in transportation of samples to off-site
 analytical facilities and the increased potential for compromise of
 sample integrity during transport, on-site analysis is an attractive
 alternative to the  conventional practice of analyzing samples in a
 remote laboratory.
   In this paper, the authors discuss their survey of existing mobile
 laboratory capabilities as they  are applicable to emergency  and
 remedial response situations. The discussion is focused on general
 design criteria, laboratory configurations, equipment and instru-
 mentation and analytical protocols that have been used in existing
 mobile laboratories. The purpose of this project was to identify im-
 portant factors that  should be  considered and evaluated when
 developing mobile laboratory capabilities for specific situations.
   A computerized search was  conducted for  articles  published
 since  1978 relating to the use of mobile laboratories for analysis
 of samples containing hazardous substances.1'6 To ensure that the
 information contained in this report was current and to describe
 areas of interest where no pertinent material was found in the liter-
 ature, the computer search was supplemented by  personal com-
 munication with experts in the field representing either commercial
 firms  or government programs. No assumptions  were  made or
 should be implied about the completeness or representativeness of
 the information so obtained.
INSTRUMENTATION

  Both the instrumental configuration and the layout within the
truck/trailer were found for various types of mobile laboratories.
These were classified as either Rapid Response Vehicles or Real-
Time Support Vehicles.
  A Rapid Response Vehicle is defined as a compact mobile lab-
oratory that can be used to determine the extent of a release, pri-
marily in air or water. A mobile laboratory designed by Ecological
Analysts, Inc. for the State of Maryland reflects this concept. The
instrumental system, which  incorporates an HP5880 gas chrom-
atograph,  was designed for air analysis but can be used to analyze
sample extracts from other matrices. Calibrated standards are pro-
duced on board by a zero-air generator and gas blending equip-
ment by Teledyne.
  Real-Time Support Vehicles are designed to assess the level of
contamination from a release and to provide field screening cap-
abilities. Commercial mobile labs as well as the facilities used by the
USEPA Municipal  Environmental Research  Laboratories,  Oil
and  Hazardous Materials Spills  Branch,  Edison,  New Jersey
(MERL-Edison or MERL-OHMSB) fall into this category.
  The commercial  mobile laboratory is typically  a  vehicle  with
appropriate support systems and a selection of sophisticated in-
struments that varies with the needs of a particular situation. A list-
ing of instrumentation used in selected commercial mobile labora-
                           Table 1
      Instruments Used in Selected Commercial Mobile Laboratories
                                    X-Ray
                                    Fluor-
                                           Cheml-
                                           1um1rt-
                                           escence  UV/VIS
                                             NOy
                                                   Spec-
  Firm*   GC  GC/MS  AA   HPLC  VOC  TOX  TOC escence Analyzer trometer


  Alert   x         x         x       x     **

  EAL    x                  xxx

  ES     x                  xx                   x

  ESE    x         xx        xx           x

  GCA    x

  ITxx^xxxxx           x       x

  OHM    x    xt    x   xtt   x   x   x                   x

  Radian  x         x   x

 VRTS    x         xx

  SCA    x         x            xx
•Acronyms refer to the following companies:
** Will be used in a pending job.
EAL = EAL Corporation, Richmond, California.
ES  = Engineering-Science, Arcadia, California.
ESE = Environmental Science and Engineering, Gainesville, Florida.
GCA   GCA Corporation, Bedford, Massachusetts.
IT   = IT Corporation, Wilmington, California.
OHM = O.H. Materials, Findlay, Ohio.
RTS = Resource Technology Services, Inc., Devon, Pennsylvania.
SCA = SCA Waste Chemical Co., Inc., Cheektowaga, New York.
tFinnigan OWA GC/MS.
ttUsage has been limited, but capability is present.
                                                                                     SCREENING TECHNIQUES      45

-------
tones, as identified in this survey, is found in Table 1. Gas chrom-
atographs (GCs) in mobile laboratories employ one or more of a
variety of detectors: flame ionization detector (FID), thermal con-
ductivity detector (TCD), photoionization detector (P1D) and elec-
tron-capture conductivity detector (HECD). Identified brands and
models of GCs used in commercial mobile laboratory applications
are the Hewlett-Packard HP5840, Perkin-Elmer P-E 3920, HNU
and Tracer Instruments Inc. Model 560. Many smaller instruments
and pieces of equipment, such  as the Miran*   IR or  bomb cal-
orimeter, are also used routinely. The U.S. Coast Guard has suc-
cessfully used a Fourier-transform infrared spectometer (FTIR) in
response to an environmental incident.
   A Canadian  firm, Sciex Ltd., has developed instruments with
mass spectrometric detector systems which can sample air directly,
and, because of the short air-sampling and sample workup times,
Sciex has begun manufacturing a mobile laboratory incorporating
them. The van is available with either a TAGA™ 3000 GC/MS
or a TAGA™ 6000 (GC) MS/MS system. Demonstrated uses of
TAGA™ systems include the following: tracing the chlorine from
a train derailment; determining PCBs in cement kiln stack gas,
ambient air and soil; air monitoring at landfills for 20 compounds
and in the workplace for bis (chloromethyl) ether; analyzing haz-
ardous waste barrel contents; direct soil surface sniffing; hazardous
waste emission monitoring; and continuous on-line monitoring of
combustion gases and automobile engine exhausts.7'9 Direct MS/
MS analysis  cannot be used alone when certain interferences are
present.' In such instances, flash gas chromatography  is used be-
fore MS/MS as a gross cleanup step.' The ionization sources for
the TAGA™ 6000 are  not completely effective in all  situations,
but the instrument has significant applications in many instances.'
 CONFIGURATION, STRUCTURE, AND SUPPORT SYSTEMS

   The effective operation of a mobile laboratory depends on its
 physical plant and support systems. Proper incorporation of such
 facilities as water supply, power, heating/ventilation/air condition-
 ing (HVAC) and related equipment is of key importance.
   In the summer,  air should be cooled approximately 5 °F below
 the desired temperature,  dehumidified and heated to the proper
 temperature to avoid furnishing makeup air at the dewpoint. A
 heating system comprised of resistance heaters of approximately
 100 amps (@ 480 volts AC) can support a 78 °F temperature differ-
 ential. A laminar flow from air diffusers improves fume  hood per-
 formance by eliminating eddies in the system.
   Mobile laboratory exhausts should be treated by passage through
 high efficiency particulate air (HEPA)  and charcoal filters prior to
 release when the unit is used  to analyze  unknown or potentially
 hazardous samples. HEPA and/or charcoal filters may be neces-
 sary to treat the intake, depending on hazards present  and loca-
 tion of the trailer. One hundred percent outside air must be used
 for all supply  and  makeup air; as many as 3 or 4 air conditioners
 may be required to furnish the 125-150 linear ft/min. face velocity
 required at the hood.
   Power is usually obtained by tapping into a utility line or indus-
 trial plant source. The power requirement is  about 100 kw (3-
 phase, 480-volt, 200-amps) for the MERL-Edison mobile labora-
 tory and about 50-100 kw for commercial mobile laboratories.
 Power from utility lines or plant sources can be run through tran-
 sient current suppressors  to minimize  surges. The mobile labora-
 tories of most commercial  firms  have auxiliary generators for
 power in either emergency or remote situations. When not other-
 wise in use, mobile laboratories are generally "supplied  with elec-
 tricity" at the  base laboratory, both to augment standard labora-
 tory facilities and to be in  a state of constant readiness for mobile
 use.
   IT Corporation  has built  mobile laboratories for the USEPA
 (IERL-CI)  and private firms and maintains  the MERL-Edison
 mobile laboratory. Each  laboratory is custom  built for a prede-
 termined purpose(s). The trailers preferred by IT are 45 ft long x 8
                                                         ft wide x 13.5 ft high, which allows inside dimensions of about
                                                         45 ft x 7.5 ft x 7 ft. This size provides sufficient space for duct-
                                                         work, insulation and hood sashes.
                                                           The preferred suspension for a trailer is add-on air bag suspen-
                                                         sion. The IT design retains metal springs so that, in case of air bag
                                                         failure, the trailer can be moved. A trailer with air bag suspen-
                                                         sion would greatly benefit from a tractor similarly equipped. IT
                                                         shock-mounts instrumentation individually, while the MERL-Edi-
                                                         son mobile  laboratory  has complete  counter surfaces shock-
                                                         mounted.
                                                           O.H.  Materials' mobile laboratories are redesigned 42-ft  box
                                                         trailers with  air-ride suspension. They contain the necessary glass-
                                                         ware and equipment to support high volume extraction and/or di-
                                                         gestions;  25  ft of chemically resistant countertop; safety features
                                                         including a shower, eyewash stations and  first-aid kits;  two  5-ft
                                                         fume hoods; storage capabilities for 250 gal of water; and six gas
                                                         cylinders  equipped  with purification systems. The water purifica-
                                                         tion system,  consisting of ion exchange and carbon beds, supplies
                                                         analytical grade water throughout the trailer.10
                                                           EAL has a 24-ft mobile laboratory with 119 ft' of effective floor
                                                         space, air conditioning and two electrical generators (4 and 6 KW).
                                                           GCA's mobile laboratory is in a custom-designed 47.5 ft x 7.5
                                                         ft trailer. The laboratory consists of three main compartments:
                                                         the hazardous  materials handling laboratory, the gas chromatog-
                                                         raphy laboratory and  a utility  room. All  initial sample handling
                                                         and aliquoting activities are conducted  at the rear  of  the trailer,
                                                         preliminary characterization and waste compatibility analyses are
                                                         conducted in the middle of the  unit and the more complex instru-
                                                         mental analyses are completed at the front of the unit.  The labor-
                                                         atory  sections  contain collapsible  tables and  workspace to hold
                                                         additional instrumentation, if needed. The utility room houses the
                                                         ventilation hood blowers and filters, the air conditioning and the
                                                         heating system and can be used for storage space, if required.
                                                           In some cases, mobile laboratories are equipped for commun-
                                                         ication by radio or telephone. MERL-Edison employs telefac-
                                                         simile to transmit data electronically.
                                                         PROCEDURES AND PRACTICES
                                                         MERL-Edison

                                                           MERL-Edison has pioneered USEPA efforts to provide rapid
                                                         on-site mobile laboratory analytical capabilities. In order to pro-
                                                         cess the potentially  large number of  samples expected at release
                                                         sites  and/or to obtain  timely  results, many of MERL-Edison's
                                                         mobile laboratory analytical procedures attempt to save time by
                                                         modifying traditional sample workup procedures (e.g.,  APHA,
                                                         ASTM, USEPA) and/or through use of more rapid instrumental
                                                         determinative steps.  The procedures at MERL-Edison are designed
                                                         to minimize losses in precision and accuracy,  to use less space and
                                                         to generate less waste.
                                                           Traditional methods  developed for lower  levels of contamina-
                                                         tion in environmental matrices frequently involve extraction, evap-
                                                         orative concentration, chromatographic cleanup and other  steps
                                                         to remove interferences and concentrate the analyte for a reliable
                                                         instrumental response.  In  spill  responses, however, the  analyte-
                                                         to-interference concentration ratio is much higher and the identity
                                                         of the spilled material is often known. Use of more rapid and di-
                                                         rect sample preparation procedures and analytical methods is pos-
                                                         sible.  For example,  MERL-Edison has developed procedures for
                                                         organics with a rapid extraction step,  such as simply spinning the
                                                         sample with the extraction solvent."'1* For certain situations, the
                                                         extraction step has been completely omitted." Streamlined clean-
                                                         up techniques and use of positive displacement micropipets to elim-
                                                         inate multiple dilutions of samples and standards have also been
                                                         adopted.13'14
                                                           MERL-Edison has not yet been  requested to perform  analyses
                                                         to be used in litigation  but has  always concentrated on site  char-
                                                         acterization and the monitoring of remedial response efforts. How-
                                                         ever, standard analytical procedures are performed in the mobile
46
SCREENING TECHNIQUES

-------
laboratory when  it is located in Edison,  New Jersey. This indi-
cates  that  the  laboratory  is capable of  generating high-quality
analyses that can be used in litigation activities, provided support
systems equivalent to those in Edison are available on-site. It is
expected that this situation applies to all mobile laboratories now
in use.
  MERL-Edison is currently assembling a manual of their tested
analytical protocols for mobile laboratory use that will be applic-
able to approximately 240  of the CERCLA hazardous substances
in air, water or soil. The instrumentation  required to perform  the
analyses  includes  GC/MS, GC/ECD or  GC/HECD,  GC/NPD,
GC/FID, spectrofluorimeters, infrared spectrophotometers, emis-
sion spectrophotometers, carbon analyzers, inductively coupled
atomic plasma emission spectrometers (ICP) and atomic absorp-
tion spectrometers (AA).  In  addition to providing field  opera-
tion guidance for the instruments listed,  each protocol provides
guidance  on sample  isolation,  fractionation,  screening  and  the
use of appropriate safety devices.
  MERL-OHMSB has  developed specific quality assurance pro-
tocols for use  in  mobile laboratory responses.'4 Several are  de-
signed to  generate method  validation  data  concurrently with
sample results  to give  the on-site coordinator "real-time feed-
back" on analytical quality.14'18 Use of a single standard to define
the linear dynamic calibration range for a large number of com-
pounds eliminates the need for extensive preanalysis instrument
calibration.13'14 A series of QA protocols that describes the  prepa-
ration of synthetic samples for method validation  has been pre-
pared. These samples make  use of media that resemble the  site
samples  and  include  detailed  instructions for  fortification of
samples with water-soluble, partially water-soluble, water-insoluble
and volatile materials.14

Commercial Mobile Laboratories

   Many commercial firms have designed their  own mobile labora-
tory sampling, sample work-up and analytical methods. Typically,
the same quality assurance procedures that are employed  in  the
main laboratory are employed in the field units.10
Cost Factors
   According to one firm which supplies both mobile and fixed-site
laboratory services, the first two factors considered in setting a
price for a mobile laboratory activation are how rapid a response
is required and how many samples per day will be processed. If
the mobile  laboratory support  effort is of sufficient magnitude,
analytical costs may be lower than at a fixed-site laboratory.  For
example, an on-site compatibility screening (PCB, oxidant,  reduc-
tant, cyanide and sulfide) of up to 200 samples per day was per-
formed at a cost of approximately $15.00 per sample. Such large
numbers of samples can help distribute the capital costs of mobil-
ization. A two-day job would have prohibitive mobilization costs,
while a stay of one to two weeks could be cost-effective. If labor-
atory trailers are located in several areas of the country, mobiliza-
tion costs are greatly reduced.
sampling protocol requires screening of all samples for radioactiv-
ity with a Geiger counter before processing them. Any material
identified as radioactive  would be  segregated and dealt with by
radiochemical  specialists. After the  screening,  all  samples are
handled in a glovebox and/or fume hood.
  Private firms are responsible for training employees in proper
sampling decontamination and hygiene procedures as well as for
providing adequate protective clothing and respirators. In the lab-
oratory, OSHA regulations as well as company-fostered safety pro-
cedures and precautions must be followed to maintain a safe work-
ing environment.
  The mobile laboratory  operated by O.H. Materials requires that
trained personnel (e.g., chemists and technicians) determine neces-
sary precautions before handling hazardous materials contained in
sample containers. In addition to specifying safety procedures to
be followed during sample collection and  analysis, O.H. Materials
also closely regulates the on-site work areas  by designating ex-
clusion (actual waste area), contamination reduction and support
(non-contaminated) zones with access to  these areas strictly con-
trolled.'0
CONCLUSIONS

  In this  survey, the authors found that mobile laboratories of
varying degrees of sophistication have been used to conduct chem-
ical analyses of environmental media. A full array of analytical in-
strumentation can be used in a mobile laboratory if the appropriate
support systems are provided. Analysts have employed both stand-
ard and specialized procedures successfully.
  Based on a limited number of past studies, mobile laboratories
can produce analytical data equivalent in quality to fixed-site facil-
ities. To routinely achieve such performance, appropriate quality
assurance, safety and support systems must be available. As a con-
sequence  of  on-site availability, analyses are completed much
sooner in a mobile laboratory than in fixed-site facilities. This com-
bination of performance capabilities and timeliness of results offer
an excellent mechanism for obtaining  analytical data during en-
vironmental responses.
ACKNOWLEDGEMENT

  Although the research described in this report has been funded
wholly or in part by the U.S. Environmental Protection Agency
through Contract Number 68-03-3050 to the Lockheed Engineering
and Management Services Company, " it has not been subjected
to Agency policy review and therefore does not necessarily reflect
the views of the Agency. Mention of trade names or commercial
products does not  constitute endorsement or recommendation for
use by the USEPA.
 STAFFING AND SAFETY

   MERL-Edison staffs the mobile laboratory with highly trained,
 junior-level personnel because they are most agreeable to extended
 travel and long term mobile-laboratory duty. In order to coordi-
 nate the on-site work, senior-level scientists at the central labora-
 tory receive raw data from the mobile laboratory via telefacsimile.
 After interpretation of the data, the central laboratory relays spe-
 cific sample analysis procedures and corrective measures back to
 the mobile laboratory." O.H. Materials staffs its mobile laboratory
 with highly trained chemists, while sampling and ambient air mon-
 itoring are performed  by trained technicians.'0 A GCA case study
 mentions a four-man crew including a chemist.
   An important part of a mobile laboratory set-up is the provision
 for safe working conditions for on-site personnel. MERL-Edison's
REFERENCES

 1. Sem, G.J., Whitby, K.T., and Sverdrup, G.M., Adv. Environ. Sci.
    Technol., 9, 1980, 55-68.
 2. Stevens, R.K., Dzubay, T.G., Shaw, R.W. Jr., McClenny, W.A.,
    Lewis, C.W., and Wilson, W.E., Environ.  Sci. and Technol.,  14,
    1980,1491-1498.
 3. Singh, H.B. et al., Measurement of Hazardous Organic Chemicals
    in the Ambient Atmosphere, NTIS, (EPA-600/3-83-002), 1983.
 4. McClelland, N.I. and Pawlowski, H.M., Water and Sewage Works
    126(4), 1979, 50-53.
 5. Donovan,  C.L. and Parker, J.G., Ind.  Waste, Proc. 14th Mid Atl.
    Conf., Ann Arbor Science, Ann Arbor, MI, 1982.
 6. Poretti, A., Light Met.,  1980, 801-13.
 7. Literature supplied by Sciex Ltd., Thornhill, Ontario, Canada.
                                                                                          SCREENING TECHNIQUES       47

-------
 8. Tanner, S., Ngo,  A., and Davidson, W.,  "Optimizing Productivity
    for the Trace Analysis of Real-Life Samples Using Flash Gas Chro-
    matography MS/MS." Extended Abstract, ASMS Conference, Boston,
    MA, 1983.
 9. French, J.B., Davidson, W.R., Reid, N.M., and Buckley, J.A., Chap-
    ter 18 in Tandem Mass Speclrometry, F.W. McLafferty, Ed., John
    Wiley and Sons, New York, NY, 1983, 353-370.
 10. Literature supplied by O.H. Materials Company, Findlay, Ohio.
 11. Frank, U., Gruenfeld, M., Losche, R., Lafornara, J., "Mobile Labor-
    atory Safety  and  Analysis Protocols Used at Abandoned Chemical
    Waste Dump Sites and Oil and  Hazardous Chemical Spills," Proc.
    National Conference on Control of Hazardous Materials  Spills,
    Louisville, KY, May 1980, 259-263.
 12. Gruenfeld, M., Environ. Sci. Techno!., 7, 1973,636-639.
 13. Gruenfeld, M., Frank, U., Remeta, D., "Rapid Methods of Chem-
    ical Analysis  Used in  Emergency Response  Mobile Laboratory Activ-
    ities", Proc.  National Conference on Management of Uncontrolled
    Hazardous Waste Sites, October 1980, 165-172.
 14. Gruenfeld, M., Frank, U.,  Remeta, D., "Specialized  Methodology
    and Quality Assurance Procedures Used Aboard Mobile Laboratories
                                                                 for the Analysis of Hazardous Wastes,"  185th National Meeting of
                                                                 the American  Chemical Society,  Mar.  1983, American Chemical
                                                                 Society: Washington, DC.
                                                              15. Frank, U. and Remeta, D., "Rapid Quantification of Hazardous
                                                                 Materials in Sediments by Synchronous Excitation Fluorescence Spec-
                                                                 troscopy," EPA  Quality Assurance  Newsletter,  7:4, USEPA, Cin-
                                                                 cinnati, OH, 1978.
                                                              16. Losche,  R., Frederick, R., and Frank.  U., "Analysis of Oil and
                                                                 PCBs in Sediments," EPA Quality Assurance Newsletter, 3, USEPA,
                                                                 Cincinnati, OLH, April 1980.
                                                              17. Frank, U. and Pernell, L.t "Synchronous Excitation Fluorescence
                                                                 Speclroscopy,"  Analytical  Quality  Control Newsletter,  No. 31
                                                                 USEPA, Cincinnati, OH, 1976.
                                                              18. Gruenfeld, M., Frank, U., Remeta, D., Losche,  R.,  "Management
                                                                 of Analytical Laboratory Support at  Uncontrolled Hazardous Waste
                                                                 Sites," Proc. National Conference on Management of Uncontrolled
                                                                 Hazardous Waste Sites, Washington, DC, Nov. 1981,96-102.
                                                              19. Engels, J.L., Kerfoot,  H.B., and Arnold, D.F.,  Survey of Mobile
                                                                 Laboratory Capabilities and Configurations; EPA  600/X-84-I70
                                                                 USEPA, Las Vegas, NV, July 1984.
48
SCREENING TECHNIQUES

-------
                 CONSTRUCTION OF  A DATA  BASE FROM
        HAZARDOUS WASTE SITE CHEMICAL ANALYSES

                                       PAUL H. FRIEDMAN, Ph.D.
                                            WILLIAM P. ECKEL
                                             DONALD P. TREES
                                           Viar and Company,  Inc.
                                             Alexandria,  Virginia
                                             BRUCE CLEMENS
                                  U.S.  Environmental Protection Agency
                                               Washington, D.C.
INTRODUCTION

  The legislative requirements of Superfund, coupled with the ex-
istence of huge amounts of sample-by-sample data for each hazar-
dous waste site, make  the creation of an Automated Data Base
necessary to efficiently extract information from the mass of data.
Section 301 of the Comprehensive Environmental Response, Com-
pensation and Liability Act of 1980 (CERCLA)1 provides that the
President submit a comprehensive report to Congress on experience
with implementation of this Act. As part of this report, USEPA
must collect and analyze data on hazardous substances  at Super-
fund sites. Two of the sources considered to be the most valuable
for assessing the presence and extent of these substances  are  the
data and information collected as part of the Hazard Ranking
System (HRS) scoring process and the collection of analytical data
from  the  Contract  Laboratory  Program (CLP).  From these
sources,  USEPA will extract the following information:
•Sites where hazardous substances were present and where their
 abundance would pose a potential or actual hazard to  health or
 the environment
•Rank of the hazardous substances according to the  frequency of
 occurrence
•Measurement of the relative contribution of hazardous substances
 to the problem of hazardous substance releases
  There  are an estimated ten million pieces of information which
could logically be included in a comprehensive data base. This
demonstrates the need for a sampling approach to the information
for the data base. Sampling has the benefit of acquiring useful data
in the earliest stages of the data base development. This, in turn,
decreases the amount of time needed for development and, conse-
quently,  decreases the cost  of the data base.


DATA BASE DEVELOPMENT
Objective of Sampling Data to Develop  a Data Base
  A constraint on the computerized data base was to select and
automate enough data to characterize hazardous waste sites from
the CLP data and not to characterize the total data available. The
CLP data contain sample results which are indicative of site com-
position.  They  also contain  quality   control  (QC)  samples,
laboratory blanks  and  field  blanks, as well  as up-gradient and
down-gradient samples. Incorporation of the QC samples into  the
data base  would  lead to  a bias in the site characteristics. This
precludes use of a random selection of samples. A random method
would not decrease relative number of samples selected which  are
not field samples.
CLP Routine Analytical Services Repository

  There are  samples from over 1,200 sites in the CLP Routine
Analytical Services (RAS) Repository. Two-hundred thirty-one of
these sites are on the NPL. RAS samples can be classified by the
matrix (soil or water) and the type of analysis (organic or metal).
The matrix was not taken into  account in selecting samples. The
RAS analytical data have not been previously automated.
  The development strategy of the CLP Data Base has precluded
any specific intended use of the data. Thus, it has tended toward
the accumulation of a true "data base." Also, the data were not
collected using a particular mathematical or physical construct or
model beyond sampling plan methodologies. Therefore, they were
not constrained by the use of a particular model. The CLP data
have the following  conditions associated with them:
•The data were collected in and  around hazardous waste sites with
 the objective of detecting hazardous waste components that were
 in or leaving the suspected site
•While the universe of chemicals may be considered, the  most
 readily identified, quantified and validated compounds detected
 are the so-called 133 organic compounds plus metals described as
 Priority Pollutants
•These data represent analytical results and not manifests, inter-
 views from the site history or on-scene monitoring
  The Automated Data Base is  comprised of results taken from a
random selection of 30% of the sites for which the CLP had col-
lected data. The selection process was constrained by the condition
that 10%  of these sites would be NPL sites.
DESCRIPTION OF ANALYTICAL DATA SOURCES

  A sample is physical evidence collected from a hazardous waste
site, the immediate environment or any related source. All samples
collected at one site during a predetermined and finite period of
time were grouped into a Case and were assigned a Case Number
by the Sample Management Office (SMO). The Case Number pro-
vides a unique identification for all relevant documentation.
Organic Analysis Data Package

  The  Organic Analysis Data Package contains at least four
packages: an Organic Sample Data Package for each sample in the
Case; an Organic QC Package (blank/duplicate/spike); an Organic
Sample Data Summary Package; and an Organic Standards Data
Package.  All reports and required documentation are identified
                                                                                  SCREENING TECHNIQUES      49

-------
with the respective SMO Case No. and associated Sample No. and
Traffic Report numbers.
  The analysis summaries in the Case Folders (described  below)
were used as the source for the Automated Data Base. The Organic
Sample Data  Summary Package includes:  organic  analysis  data
sheet(s); tabulated results of analyses of specific compounds re-
quired by the contract and tentatively identified compounds (TIDs)
from the Organic Sample Data Package;  and  surrogate spike,
matrix  spike,  duplicate and blank analyses  from the Organic QC
Package.  This summary provides the organic analysis data  for the
analytical data base and is filed in the readily accessible SMO Case
Folder.
Inorganic Analysis Data Package
   This data package includes, for each Case, the tabulated results
of inorganic analyses. The data package includes:  the analytical
results  for waste and extract spikes,  duplicates, standards and
blanks; instrument calibration data;  and procedural/method blank
results.
Case Folder

   A Case Folder represents the sample-by-sample results for a site
over  a finite  period of time. The Case Folders  are organized  in
ascending Case Number sequence.   A Case Folder contains  the
 following:
 •Traffic Report forms for each organic,  inorganic and/or  high
  hazard sample in the Case
 •Tabulated analytical results for  all  organic  and metals determina-
  tions
 •All other relevant sampling documentation and  correspondence
 ANALYTE OCCURRENCE

   The number of samples needed to detect in the data base whether
 a pollutant was found at a site is a  function of the number  of
 positive samples for that pollutant at the site. In the following rela-
 tionship:2
         F =  1 - (1 - P)<>
                                                  (1)
   P is the proportion of positive samples from the site and F is the
 probability of getting at least one positive sample out of n random-
 ly selected samples. Six samples selected randomly  from the CLP
 data for a site would be sufficient to detect substances at sites where
 at  least half the  samples  were positive for a given hazardous
 substance. There  is a 98% chance that  at least one of the six
 samples would show a positive result. This  probability, combined
 with resource constraints, led to the inclusion of six determinations
 of each pollutant at each site.
   Possible false positive results point to the need to establish a
 minimum concentration below which the presence of the pollutant
 is not convincingly demonstrated. Before a sample may be con-
 sidered positive from a hazard standpoint,  the significance of oc-
 currence of compounds at  or  below specified  indices (i.e., water
 quality limits) may need determination. The  number of samples (n)
 needed to estimate the average concentration at a  site with 95%
 chance of making an error no greater than E% of the true value is:1
         n = 4 CVE2
                                                  (2)
where C is a measure of random  between-sample variation as a
percentage of the true concentration expressed as the coefficient of
variation for the site. The values of n generated for different values
of E and C2 are given in Table 1. For example, six organic samples
may be sufficient if a 50% estimation error in concentration is ac-
ceptable for a site.

DATA ENTRY  PROCESS

  The data entry process is initiated when a completed C'ase Folder
is received. The Case Folders are  first reviewed by a pre-entry
                                                         editor. This edit is verified and the Case Folder, if ready for entry,
                                                         is given to a Data Entry Operator for further processing. A post-
                                                         entry editor checks for errors. When all errors have been rectified,
                                                         the information  is entered  into the system and merged into the
                                                         Master File.

                                                         Pre-Entry Edit and Validation

                                                           The pre-entry editor  initiates the data entry  process. The pre-
                                                         entry editor: (1) checks the physical contents of the Case Folder for
                                                         completeness; (2) correctly identifies, names and lists all samples in
                                                         the Case to be included; (3)  locates and identifies tentatively iden-
                                                         tified compounds and flags all compounds  not identified in our
                                                         Chemical Compound Nomenclature List (CCNL) for resolution or
                                                         updating; (4)  records all  relevant administrative data  (sampling,
                                                         analysis dates, etc.);  (5)  selects a  matrix code;  (6) reviews the
                                                         analytical data for each sample; and (7) flags and resolves any data
                                                         reporting problems. If  analytical and/or administrative data are
                                                         missing,  the  folder is retained until the missing  information is
                                                         located.

                                                                                    Tibk I
                                                                         Values of n dcneraled by Eq. 2
Coefficient of
Variation (C)
25
30
50
60
Percent
Error (E)
25 50
4
6
16
23
1
2
4
6
  If a compound cannot be identified, it is noted and filed on an
Unknown Compound List. Provisions are being made to enter such
compounds  in general  categories  under the rules  set  by the
Chemical Abstracts Service. The Chemical Nomenclature File cur-
rently holds entries for approximately 1,274 organic and inorganic
substances.
Sample Selection

  Samples were selected systematically from a Case, excluding QC
samples, with the following assumptions:
•For sites  having data in more than one Case, the first Case con-
 tains the most representative samples and results  of site consti-
 tuents
•The general usefulness of the  Automated Data Base will be en-
 hanced by including data from a larger number of sites
  The data sampling procedure consisted of systematically select-
ing a fixed number of samples from 30%  of CLP sites. If a selected
number of sites in the  Automated Data Base exhibited a  great
amount of variation  from the total data for this site, additional
samples would be added on a site-by-site basis after the initial stage
of data base development. Program resources allowed  for automa-
tion of the results of  10  samples for each of the 358 sites selected.
Organic Compounds

  The samples within a Case are aligned in ascending  numerical
order. The first and  last samples are selected as well  as every nth
sample where n is a multiple of the N/5th sample (N being the total
number of samples in the Case). When a QC sample is  selected, the
lower adjacent sample is selected in its place. If the first Case does
not contain enough samples, the second Case is sampled using the
same method, the nth sample being the multiple of N/(k-l) where k
eguals the number remaining to be sampled.
Metals

  The method of selection is identical to that of organic samples
except that k for the  first Case is 3.
50
SCREENING TECHNIQUES

-------
Verification Rationale

  The approximation of a model requiring random  selection of
samples (Eq. 1 and Eq. 2) to a systematic selection required verify-
ing the correspondence of the Automated Data Base with the CLP
Repository from which the data was drawn.
  A sample requires fewer resources to assemble a data base but
runs a greater risk of not adequately representing the data on a site
basis. Top large a sample would assuredly give a good representa-
tion but it would unnecessarily tax resources and necessitate sam-
pling a smaller number of sites. Comparing the CLP data and the
site data included in the Automated Data Base gives a qualitative
evaluation of the correspondence between the two distributions.
The criterion for correspondence of the data base with the CLP
data was the correlation of the site-by-site frequency distribution of
compounds  in  CLP data  to the frequency  distribution of com-
pounds in the Automated  Data Base.
  The degree of correspondence of the Automated Data Base to
the site inventory  must be quantified so that the sampling pro-
cedure can be validated. The distribution of compounds obtained
by sampling the CLP data on a site-by-site basis should be propor-
tional to the distribution that  exists  in the CLP  Repository. The
correlation coefficient  of the linear function defined  by the com-
parison of the frequency distribution of the CLP inventory with the
data base sample distribution should be a quantitative measure of
the degree to which the data base characterizes the site data. It is,
therefore, an indirect evaluation of the sampling procedure. If the
samples in the data base exactly replicated the distribution in the
CLP site inventory, the line described by the points would be a
straight line of  slope, M = 1, through the origin. Computation of
the  correlation  coefficient  quantifies  the  degree to which the
Automated  Data Base describes the data inventory for the site in
the CLP data.

RESULTS

   Histograms of the frequency of compound detection for a par-
ticular site are shown in Figure 1. The lower histogram is the fre-
quency distribution of detection for all the CLP data for the site.
The upper histogram is the distribution for the results incorporated
into the data base.
  A typical plot of the frequency distribution of the percent abun-
dance of the Automated Data Base versus the percent abundance in
the CLP Data Base for a particular site is given in Figure 2. Each
point in Figure 2 represents the percent of samples that are positive
for a particular compound. The  X coordinate represents the per-
cent in  the CLP site inventory and the Y coordinate represents the
percent of positive results from the sample taken for the data base.
Positive results  are restricted to certain percentages as a function of
the number of samples taken; i.e., for six samples the only possible
percentages are 0, 17, 33,  50, 67, 83 and 100.
                                  hCh
LL
                                   KTTl
                             Figure 1
   Frequency Distribution of Compound Abundance in a Site Inventory for
       VOA Data (Bottom) and Corresponding Distribution'in the Data
                            Base (Top)
                  BASE
                        100-1
                         75-
                         50-
                         25-
                                      25
                                 50
75
                                                                    100
                           SITE  INVENTORY
                           Figure 2
                     Automated Data Base
               Site Data vs. CLP Site Data (Volatiles)


  Tables 2 to 5 contain the correlation coefficients obtained by the
sampling procedure described and by introducing modifications to
the sampling procedures as described below:
•CLP refers to the described systematic procedures
•Random refers to random selection of samples using fixed num-
 bers (six organic and four metal samples)
•Proportional by Case/Random by Sample refers to weighting the
 number of samples selected from each Case as a fraction of the
 relative number of samples from each Case. The conditions indi-
 cate results for a fixed number of samples and fixed percentages of
 total samples
•Proportional by Matrix/Random  by Sample, using fixed number
 and fixed  percentage, refers to a random selection of samples
 from each matrix regardless of Case. Samples are selected in num-
 bers proportional to their matrix representation in the site data.
•Augmented by Case refers to addition of six organic samples and
 four inorganic samples proportionately by Case to the described
 systematic procedure; used  where original  single  Case  sampling
 had been done on multi-Case sites.
  Examination of the data in Tables 2 to 5 verifies that the pro-
posed sampling procedure is adequate to obtain representative data
for sites of three Cases (sampling episodes) or less. The correlation
coefficient between the Automated Data Base and the CLP site in
ventory becomes less positive when more  than three  sampling
episodes occur at a site. Use of a Case-weighted sampling improves
the correlation between the CLP Repository and the Automated
Data Base.  The great majority of sites exist in the CLP Repository
as less than three Cases,  and therefore, the systematic  sampling
scheme from the initial Case will suffice for the bulk of the CLP
data sampled.
  As a practical matter, the correlation coefficient of total data to
sampled data from each site cannot be determined. This  would be
equivalent to entering all data into the CLP Automated Data Base.
It  was necessary to  augment the sampling of certain sites  where
more than  three Cases of data were in the CLP Repository. The
correlation  coefficients for  the  augmented  data  to  previous
methods are compared in Tables 6 and 7. Overall correlation to the
total site data was improved over the single Case sampling method.
This augmented method helped to maintain consistency  of opera-
tion  with the original sampling scheme.
   It  is interesting to speculate on the physical interpretation of the
intercept and the slope of the line defined by the frequency distribu-
tions of the site inventory and the Automated Data  Base sample.
The  intercept would be non-zero if the X or Y coordinate of each
point had a constant percentage of positive values added  to or sub-
                                                                                         SCREENING TECHNIQUES
                                                                        51

-------
                            Table 2
     Site Data Characterization for Small (less than 3 Cases) Sites
                       Organic Compounds


Site
Wildcat Landfill
Toms River
Mason County
Marion City
No, of
Data
Points
18
42
27
14

Correlation
Coefficients
0.79
0.91
0.86
0.86


Slope
1.62
1.12
0.91
0.90


Intercept
-7.24
3.06
3.29
-2.25
Mean/Median Correlation Coefficients
Organic - 0.86/0.86
                            Table 3
      Site Data Characterization for Small (less than 3 Cases) Sites
                             Metals


Site
Wildcat Landfill
Mason County
Marion City
No. of
Data
Points
18
17
11

Correlation
Coefficients
0.92
0.58
0.73


Slope
0.85
0.94
1.31


Intercept
18.54
-9.36
-5.64
  Mean/Median Correlation Coefficients
  Inorganic - 0.74/0.73

                             Table 4
     Site Data Characterization by Single Case (Old) Sampling and
                    Multi-Case (New) Sampling
                       Organic Compounds

No.
of
Site Cases
Old Mill
Indian Bend Wash
Envirochem
Andover Sites
Acme Solvents
Niagara Frontier
Kingston, NH
4
6
5
4
4
4
}
No. of
Data
Points
73
20
52
24
56
39
68
Correlation
Coefficients
(Old/New)
0.
0.
0.
0.
0.
0.
0.
73/0.
73/0.
52/0.
53/0.
76/0.
91/0.
69/0.
81
97
61
87
90
88
89
Slope
(Old/New)
0.
0.
1.
1.
0.
1.
0.
76/1
49/1
12/0
24/1
68/1
63/0
95/1
.06
.30
.71
68
.2}
.86
.06
Intercept
(Old/New)
-4.07/-0
-3.31/-2
12.7/0
1.66/-4
9. 39/0
-4.24/-2
-2.00/-1
.73
.79
.75
.18
.67
.64
.70
 Mean (Old/New) - 0.70/0.85
 Median (Old/New) - 0.73/0.88
                            Table 5
     Site Data Characterization by Single Case (Old) Sampling and
                    Multi-Case (New) Sampling
                             Metals

No.
of
Site cases
Old Mill
Indian Bend Wash
Envirochem
Andover Sites
Acme Solvents
Niagara Frontier
Kingston, NH
3
6
3
3
2
2
4
No. of
Data
Points
22
20
21
16
12
16
21
Correlation
Coefficients
(Old/New)
0.
0.
0.
0.
0.
0.
0.
84/0.
49/0.
50/0.
72/0.
97/0.
97/0.
42/0.
90
36
88
95
92
95
89
Slope
(Old/New)
1.
0.
0.
1.
1.
1.
0.
22/0.
81/0.
64/1.
02/1.
02/0.
13/0.
55/1.
82
57
05
00
92
91
15
htercept
(Old/New)
-15
-7
.7/-9.
.18/3.
22.9/7.
-0.
-3.
-I.
30
74/-9.
84/-0.
16/-12
.0/-8.
06
40
22
02
68
.5
S3
                            Table 6
          Comparison of Correlation Coefficients for Large
                        Multi-Case Site*
                            Organic
Site
Old Mill
Indian Bend Wash
Envirochem
Andover Sites
Acme Solvents
Niagara Frontier
Kingston, NH
Mean
Median
Original
Systematic
0.73
0.73
0.52
0.53
0.76
0.91
0.69
0.70
0.73
Proportional
by Case/
Random by
Sample
0.84
0.97
0.61
0.87
0.90
0.88
0.89
0.85
0.88
Augmented
Proportional
by Caw
0.89
0.83
0.76
0.78
0.93
0.90
0.78
0.84
0.83
                                                                                                  Table?
                                                                                  Comparison of Correlation Coefficients for
                                                                                           Large Multi-Cue Sites
                                                                                                 Inorganic
Site
                     Original
                     Systematic
               Proportional
               by Case/
               Random by
               Sample
                Augmented
                Proportional
                by Case
Old Mill
Indian Bend Wash
Envirochem
Andover Sites
Acme Solvents
Niagara Frontier
Kingston, NH
                                                                     Mean
                                                                     Median
0.84
0.49
0.50
0.72
0.97
0.97
0.42
0.90
0.36
0.88
0.95
0.92
0.95
0.89
0.91
0.84
0.84
0.87
0.98
0.97
0.64
                     0.70
                     0.72
               0.84
               0.90
                0.87
                0.86
Mean (Old/New) - 0.70/0.84
Median (Old/New) - 0.72/0.90
traded from it. This would indicate a bias in the subset of samples
included in the  Automated Data Base. The slope of a subset of
samples which has a frequency distribution identical to the CLP
site inventory would be equal to 1. To the extent that the  slope
changes from 1, the characterization becomes a proportionality.
Slopes less than  1 may indicate an under-representation of positive
  The correlation of the frequency distribution of positive occur-
rences with the frequency distribution obtained from a systematic
sampling of the first Case may have some inferences with regard to
the cost/benefit of subsequent sampling of hazardous waste sites.
Frequency distributions indicate  that the  sites are  no  better
characterized with  respect to compounds present in two  or three
Cases  than in the  first collection of samples. This conclusion
disregards other objectives of subsequent site visits in connection
with sampling such as determining the extent of contamination.

ACKNOWLEDGEMENTS

  This work  was funded by the U.S. Environmental Protection
Agency   Office  of  Emergency  and   Remedial  Response,
Washington,  D.C., under  USEPA Contract No. 68-01-6702 and
No. 68-03-3113. The authors wish to acknowledge the assistance of
Ms. Pat Murray and Dr. Kaveh Sotoudeh for materials provided
and Ms. Barbara Jean for manuscript preparation.

REFERENCES

1.  Public Law 95-10, 94 Stat.  2767 (1980), Codified at 42 USC, Sec. 9601
   et. seq.
2.  Elder, R., Food  Safety Division,  U.S.  Department of Agriculture,
   Private  Communication.
52       SCREENING TECHNIQUES

-------
                      APPLICATION  OF  MOBILE  MS/MS  TO
                HAZARDOUS  WASTE  SITE  INVESTIGATION

                                            DAVID BEN-HUR, Ph.D.
                                            JAMES S. SMITH, Ph.D.
                                               Roy F. Weston,  Inc.
                                           West Chester, Pennsylvania
                                              MICHAEL J. URBAN
                                     U.S. Environmental Protection Agency
                                            Emergency Response Team
                                                Edison, New Jersey
INTRODUCTION

  In recent years, application of the MS/MS technology has been
advanced significantly in the areas of air monitoring and dioxin
analysis. In both applications, the technique offers advantages that
are unparalleled by conventional modes. Specifically, the technique
offers real-time analysis; detection, identification and quantifica-
tion at very low levels; and high specificity. Results of both types of
applications are discussed in this paper. This discussion, however,
is preceded by some details of the instrumentation.
INSTRUMENTATION

  MS/MS has recently emerged as an analytical tool. The instru-
ment consists of two quadrupole mass analyzers separated by a
third quadrupole operated in a total ion mode.  By introducing a
collision gas in  the domain of the middle quadrupole, the ions
emerging from the first mass analyzer are fragmented, and the
fragments  are analyzed in the second mass analyzer. The instru-
ment can be operated in various modes. In a single mass analyzer
mode, the instrument is  used to scan all ions  produced by the
source (parent ions). In the tandem MS mode, the instrument can
be set to transmit preselected parent ions  through the  first mass
analyzer, fragment the selected parent ion in the region of the mid-
dle quadrupole and analyze the produced fragments in the second
mass analyzer, thus producing characteristic mass spectra that are
used in the identification of components. Alternatively, the second
mass analyzer may be set to  monitor  a  specific fragment ion
(daughter ion). In the last two applications, the first mass analyzer
acts  as  a mass  separator,  thus  eliminating the  need  for
chromatographic separation.
  Because  the fragmentation pattern of  molecules  is a unique
characteristic, the combination of specific parent ion and daughter
ion offers high specificity in the analysis. When this combination is
not sufficiently unique, monitoring one parent ion in  tandem with
two daughter ions supplies the additional specificity.
  Two  ionization sources  are  currently  in use.  Both  employ
chemical ionization. This mechanism of ion formation is such that
the parent  ions formed are more readily related to the compound
being analyzed than is the case when electron impact  is  used. The
dominant mechanisms of ion formation are charge  transfer and
proton transfer,  but with little or no fragmentation of the parent
ions at the  source. Hence, the molecular weight of the  neutral com-
pound is readily  determined when chemical  ionization  is employed.
The two sources  that are employed in the field are described below.
The Atmospheric Pressure Chemical
Ionization (APCI) Source

  In the APCI source, the principal components of air are ionized
by a corona discharge, forming initially the ions Nj"and Of These
energetic ions rapidly undergo collisional moderation and through
reaction and charge transfer form hydrated protons H + -(H2O)n,
where n is a whole number. The hydrated protons, in  turn, are
responsible for the ionization of the trace components. This ioniza-
tion usually is in the form of proton transfer, so that the parent ion
is in the form TH +, displaying an apparent molecular weight 1 amu
larger than the true molecular weight. In addition to the parent ion
of the form TH+, hydrated parent ions are also formed. These ions
are  of the form TH+ -n(H2O), where n is a whole number. The ap-
pearance of clustering is a function of the moisture content of the
sample. In spite of the appearance of more than one parent ion,
however, the recognition of clusters is made easy by virtue of their
exhibiting a series of daughter ions that are formed by a neutral loss
of 18 amu, corresponding to the sequential loss of water molecules.
  Trace compounds that are amenable to ionization by this techni-
que are  those with high proton  affinity, characteristically  ox-
ygenated and nitrogenated compounds. Selectivity in the ionization
mechanism is obtained by introducing into the air stream a reagent
gas, such as ammonia, which has a proton affinity higher than that
of water. If ammonia is used as a reagent gas,  the principal ioniza-
tion source will be the ammonium ion, NH$,  and trace amines in
the  air stream will be preferentially ionized without interference
from the oxygenated compounds.
The Chemical Ionization (CI) Source
  Initial ionization in the Cl source is similar to that which occurs
in the APCI source, forming the ions N| and Oj. The Cl source,
however, is operated at reduced pressure, typically 0.3 torr.  Thus,
collisional moderation of the very energetic ions is reduced, and the
principal ionizing medium is NO+, which is formed by reaction be-
tween the initial ions and neutral oxygen and nitrogen. Ionization
of trace components occurs through several mechanisms principally
charge transfer, forming a parent ion T+ with a mass equal  to the
molecular weight of  the neutral trace  compound  or by proton
abstraction forming the ion (T - H) + with a mass 1 amu less than
that of the neutral parent compound. Additionally, the presence
of moisture in the air would lead to ionization reactions similar
to those occurring in the APCI source.  Clustering leads to parent
ions of the form T'NO+. In the case of chlorinated compounds,
loss of chlorine or hydrogen  chloride may  occur at the source.
                                                                                    SCREENING TECHNIQUES      53

-------
                                                           Figure 1
                                     A Schematic Representation of THAT TAGA™ 6000 MS/MS
While the complexity of the source chemistry complicates the in-
terpretation of the results, the  Cl source provides a means for
direct analysis of aromatic compounds, the alkenes and chlorinated
compounds.
  The design of the instrument is shown in Figure 1.


APPLICATIONS TO AMBIENT AIR ANALYSIS

  The existing conventional methodologies for the detection of low
levels or organics in ambient air require that the compounds be ad-
sorbed onto a substrate over a finite period of time. The substrate is
then taken for analysis in an off-site laboratory where  the com-
pounds are either thermally desorbed or solvent eluted  from the
substrate and analyzed by GC or by GC/MS. The technique suffers
from several drawbacks:
•The process is time consuming, producing results several days or
 weeks  after the sampling  has been performed.
•Obtained quantitative results are always time-averaged over the
 duration of the sampling period. Temporal fluctuations cannot be
 obtained by this method.
•The desorption process for the determination of the components
 collected  introduces  uncertainties.  Certain compounds are ad-
 sorbed so strongly that they cannot be recovered from the sub-
 strate.  Other compounds may undergo reactions or rearrange-
 ments so that they are misidentified in the recovered eluate.
  These difficulties are minimized or eliminated by employing the
on-site MS/MS instrument mounted  in a van and operable both in
a stationary mode and a mobile mode. The specific advantages are:
•Analysis is in real time.
•Quantitation is performed so that temporal fluctuations  are ob-
 tained, yet time-weighted average concentrations can also be de-
 rived from the data.
•Alteration of the components of air is minimized.
•Wall effects are reduced by maintaining a very rapid air flow
 through the  system, typically 21/sec.
•The instrument is capable of performing analyses of extremely
 polar compounds.
  The instrument has been employed in the mobile mode  in several
studies.  Two of these studies are presented here.

Abandoned Hazardous Waste Sites

  In the first study, the mobile laboratory was called upon to
qualitatively determine trace components in  the air and quan-
titatively determine specific target compounds. The site under con-
sideration had been abandoned  by its owners, closed, and at the
time of  the study was in the process  of being cleaned up. The site
contained an incinerator,  several ash piles, waste lagoons and
drums.
                                                          Several areas of the abandoned site were investigated qualitative-
                                                        ly using both the APCI source and the CI source. Without attemp-
                                                        ting to identify components, a single MS scan was taken off-site
                                                        and upwind of the site so that the total ion spectrum of the ambient
                                                        air  could be obtained. The total ion spectrum  was  used as  a
                                                        reference. A total  ion  spectrum was taken repeatedly at  various
                                                        locations on-site. In  each case,  the reference scan was subtracted
                                                        from the new scan, and the masses of components that were absent
                                                        in the  background were subjected to MS/MS analysis. The first
                                                        mass analyzer was set to transmit one parent ion at a time, and the
                                                        second mass analyzer was set to  analyze and record the fragmenta-
                                                        tion pattern. The background subtracted total ion scan is shown in
                                                        Figure 2 while the resulting mass spectrum of a single component  is
                                                        shown in Figure 3. The compounds that were identified at this site
                                                        are listed in Table 2.
                                                                                           l«0  160  ISO  ZOO  Z20   fX
                                                                                   Figure 2
                                                          Background Subtracted Single MS Scan of Air Above a Sludge Pile
                                                           Quantitation was performed for specific target compounds. To
                                                         establish  method  equivalency,  the  MS/MS  technology  was
                                                         employed side-by-side with conventional NIOSH charcoal  tube
                                                         sampling. A typical single component calibration curve is shown in
                                                         Figure 4. Because quantitation was to be performed alongside char-
                                                         coal tube sampling, the mobile laboratory was stationed in one
                                                         location.
                                                           Activity on the site, however, created fluctuations in the concen-
                                                         tration of the target compounds. The effect is shown in Figure 5. A
                                                         summary of the time-weighted average concentrations of the target
                                                         compounds obtained by the two methods is given in Table 2. The
                                                         data clearly  demonstrate the equivalency  of the two  methods;
54
SCREENING TECHNIQUES

-------
lea-

 9O-
 79-

 te-
                                            PCPK-BOCXGBCmD
                    .1.  I.
           »••   «•"    sa.a    ea'.e   TB'.B   ee'.a  ' se'.e  w.

                         Figure 3
      Normalized MS-MS Scan of M/Z = 97 from either
             Trichloroethane or Dichloroethylene
                           Figure 5
         Time Dependent Measurements of the Concentration
                    of 1,1,1-Trichloroethane
                          Table 1
Compounds Identified at Abandoned Site Using MS/MS Technology
                            Table 2
      Comparative Results from MS/MS and Conventional Analyses
Compound
Fornamide
Acetone
Acetanlde
Propanol
Ethylene glycol
Methyl ethyl ketone
N.N-Dinethylfornanide
Butanol
Dinethyliultoilde
Pyrldlne
Hethylene chloride
or chlorofprm
N,N-Dlnethylacetanide
Aninobutanol
Toluene
Aniline
Dichloroethylene
or trichloroethane
Methyl isobutyl ketone
N-Hydroxy-l,2-ethylene-
diaaine
Diethylene glycol
Xylene
Chlrobenznene
N,N-Diethylacetanide
butyl Cellosolve
Tr ichlor oethylene
Tr ichlor of luorome thane
Dichlorobenzene
Tetrachloroethylene
Source
APC1
APC1
APC1
APC1
A PCI
APC1
APC1
APC1
A PCI
APC1
Cl

A PCI
APC1
APC1
APC1
Cl

APC1

APC1
APC1
Cl
Cl
APC1
APC1
Cl
Cl
Cl
Cl
Molecular
Weight
45
56
59
CO
62
72
73
74
76
79
84
lie
87
89
92
93
96
132
100

104
106
106
112
115
lie
130
136
146
164
Parent
Ion m/t
46
59
60
61
63
73
73
75
79
80
63
63
68
90
91
94
96
96
101

105
1CT7
106
112
116
119
130
101
146
164
   zeeoa.
   loxe.
                 s.e      IO.B      is.e
                   CONCCWTRflTION (PFrlt
Compound
Methylene chloride
1 , 1 ,1-Tr ichlor oe thane
Tr ichlor oethylene
Tetrachloroethylene
Chlorobenzene
Toluene
Xylene
Methyl ieobutyl ketone
Acetone
Mean Concentration,
ppm ( v : v )
MS/MS Method NIOSH Method
3.4
1.4
1.8
0.2
0.04 Not
0.7
0.5
0.005 Not
0.017 Not
3.7
1.7
1.1
0.25
detected
0.5
0.29
detected
detected
                          Figure 4
 Calibration of 1,1,1-Trichloroethane Parent/Daughter Ion = 99/61
however, the results from the MS/MS quantitation were available
on the same day, while the results from the charcoal tube quantita-
tion became available two months after the samples were collected.
Ambient Air Analysis in the Vicinity of
a Sewage Treatment Plant
  In this study, attempts were made to identify disagreeable odors
that pervade a residential area. The location of the study is a heavi-
ly industrial area, interspersed with residential sections.
  Odor incidents have been reported under certain meteorological
conditions, usually in the summer. Previous studies and inspections
have indicated that the  odor is very intense in the vicinity of the
sewage treatment plant (STP). The goal of this study was to iden-
tify the odorous components and to attempt to  isolate what in-
dustrial clients of the  STP  might be contributing to the  odor
problem.
  During  the period  of the  study, there was no odor incident,
although the odor was very intense in the immediate  vicinity of the
treatment plant. In this study,  all measurements  were  done from
public access roads.
  The industrial nature of the area and the constantly  shifting
winds  made it  difficult  to  determine a baseline  background.
Because of the industrial background, the reference scan  for this
study was obtained with "zero air" distilled from liquified air and
presumably containing  only  the principal components of air.  A
large  number of compounds was identified, although it is not
known whether any of  them contributed to  the  odor. The com-
pounds identified  and a notation showing whether they appear
downwind  of industries that may be contributing  through their
                                                                                         SCREENING TECHNIQUES      55

-------
                            Table 3
                Summary of Identified Compounds
                 Moleculer   STP     STP    Ind.  Ind.  Ind.
      Compound       Height  Upwind  Downwind   ABC
AAMonl*
H«th«nol
Hydrogen eulflde
Acetonltrlle
Ethtnol
Acetone
17
32
34
42
46
56
X
X
X

X
X
X
X
X
X
X
X
X
X


X
X
X
X

X
X
X





X
      PropylAnlne      59
      Pcopanol         60
      DUethyl «ulflde  (2
      Ethinethlol      62
      Ethylene glycol
      Methyl ethyl
       ketone
      Buttnol
      Propenediol
      Benzene
      Methylen
       chlorid
      Pyrrolld
      ThUiole
      Methyl •
      Butyric
rylate
eld
      Toluene
      ftethoiyCur an
      Hethylthlnole
      Nethyl leobutyl
       ketone
      Propyl acetete

      Hexenol
      Xylene
      Dlethylene
      9lycol
      Chlorobenlene
      Octtnol
       (2

       72
       74
       76
       78
 84
 85
 85
 86
 86

 92
 98
 99

100
102

102
106

106
112
130
      Dlchlorobentene  146
waste to the plume emanating from the STP are shown in Table 3.
Although  a strong correlation between the composition found at
the STP and that found at Industry B appears, additional analyses
are needed before a stronger tic can be established.
  Because of the possibility of contribution of compounds from
several sources, mobile monitoring for target compounds was per-
formed. The principal target compounds were benzene, toluene
and dichlorobenzene. Around the STP, the compounds peaked
simultaneously (Figure 6). The peaking at Scan No. 167 occurred
exactly downwind of the STP, with no potential sources between
the van and the STP.
  A similar mobile monitoring was performed around Industry B
(Figure 7). Simultaneous peaking is observed at Scan No. 140,  in-
dicating that at  least  in part the compounds of interest  occur
simultaneously. However, additional sources of the compounds are
also indicated.
  Concentrations and detection limits  for several  target  com-
pounds were determined (Table 4).  The  reported detection limits
are as measured in the field. Lower detection limits can be obtained
if the instrument  is optimized for the specific target compound.
                                                    APPLICATION OF MS/MS TO DIOXIN ANALYSIS

                                                      In the past two years, considerable effort has been expended to
                                                    develop the application of MS/MS technology to dioxin analysis.
                                                    To facilitate efficient use of the instrument, both sample prepara-
                                                    tion  and  chromatographic separation  had to be modified drasti-
                                                    cally. Because of the high specificity of the MS/MS technique, it was
                                                    hoped that the extensive sample preparation  that is currently re-
                                                    quired by the USEPA method could be reduced to much simpler
                                                    procedures. The high sensitivity of the instrument made it possible
                                                    to perform the analyses on extracts without the need for preconcen-
                                                    trating.
                                                      The method consists of single-step extraction, extract clean-up
                                                    when needed, flash  chromatography  and MS/MS  detection. A
                                                    schematic diagram of the method is shown in Figure 8.
                                                      Experience with Missouri soils indicated that sample extracts
                                                    rarely require cleanup. New Jersey soils, on the other hand, almost
                                                    always need to be cleaned up.
                                                      Tetrachlorodibenzodioxin (TCDD) is introduced into the instru-
                                                    ment via a capillary GC column programmed at 20°C/min. Under
                                                    these conditions,  the peak  elutes in about 5  min. In  the source,
                                                    TCDD is  ionized to give a parent  ion of the same  mass as the
                                                    molecular weight of the neutral compound:
                                                                          TCDD + NO
                                                                                   •TCDD+ + NO
                           Figure 6
      Mobile Monitoring for Benzene, Toluene, Chloroform and
          Dichlorobenzene around Sewage Treatment Plant
                      ^"—•ww-vJ \~+i*~*.
                            Figure 7
     Mobile Monitoring for Benzene and Toluene around Industry B
                                                                                              Table 4
                                                                        Detection Limits and Maximum Measured Concentrations of
                                                                                  Selected Compounds In Ambient Air
Concentration, ppb
Compound
Benzene
Toluene
Dichlorobenzene
Chloroform
Trlchloroethylene
Acetone
Methyl ethyl ketone
Hethylthlazole
Detection
Limit
J
2
0.3
4
30
IS
10
1
STP
60
30
12
ISO
ND
70
14
14
Has i ma*
Induatry B
7
80
3
NO
ND
30
ss
11
56      SCREENING TECHNIQUES

-------
                I Weigh vial containing about 5 g. anhydrous sodium sullala |
                \ Add about 5 g. solid sample, and reweigh
                                                        I
                         '
                | Add internal standard and surrogate compounds
                         I
                I Add 5 ml. solvent mixture, shake lor t minute
                '         I
                [ Centrifuge tor 2 minutes
                        I
                  With syringe, withdraw 1 ml. of extract Force through
                  0.5 um Teflon filter into 9 ml. of distilled water
                 Shake water-extract mixture, centrifuge and withdraw
                 the insoluble bubble
                I ANALYZE BY GC-MS-MS I
                                      No
                                                _| Clean extract through Procedure A |
                                                         J_
                                                  ANALYZE BY GC-MS-MS
                | Calculate Quantity I
                                                                                  J Clean extract through Procedure D |
                                                                                   | ANALYZE BY GC-MS-MS |
                                                  I Calculate Quantity]
                   OPTION X
                                                      OPTION XA
                                                                                   | Calculate Quantity

                                                                                       OPTION XAD
                                                             Figure 8
                                   TCDD Sample Preparation and Analysis Scheme Using GC/MS/MS
Typical fragmentation of TCDD + is shown in Figure 9. The major
ions formed during fragmentation are:
  m/z
Identity
320       TCDD +, parent ion
285       (TCDD - Cl) +, neutral loss of chlorine from parent ion
257       (TCDD - COC1) + , neutral loss of COC1 group from
            parent ion
222       (TCDD - COC12) + , neutral loss of COC1 and Cl from
            parent ion
194       (TCDD - 2COC1) + , neutral loss of two COC1 groups
            from parent ion
                                                   233  308 323
                             Figure 9
                  Fragmentation Pattern for TCDD
                         Parent M/Z = 320
  Instrumental parameters are optimized so that the parent ion is
minimized and the fragment ion of mass 25" is maximized. Under
these conditions, native TCDD will also exhibit a fragment of mass
259 due to the natural distribution of chlorine isotopes. The sur-
rogate, 37C14-TCDD, will appear at mass 263, and the internal
standard, 13c12-TCDD, will appear at mrjs 268.  Hence, in the
analysis, the instrument is set to monitor four parent-daughter ion
pairs. These are shown below:
Component
Native TCDD
Parei ion
m/z
320
Daughter Ion
m/z
257
                                                                     Surrogate, 37ci4-TCDD
                                                                     Internal standard
                                                                                      322
                                                                                      328
                                                                                      332
                                                       259
                                                       263
                                                       268
                                                 Because this capability of the dual selection of ions provides for
                                               the selectivity  of  the  method,  the  extensive sample clean-up to
                                               remove interferents is  not necessary.
                                                 The reproducibility  of the system has been verified in an actual
                                               field application by analyzing a 10 ppb standard on a daily basis.
                                               The results  are  shown in  Table  5.  Another measure  of the
                                               reproducibility of the technique was obtained by repetitive analyses
                                               of a well-homogenized soil. These results are shown in Table 6.
                                                 More recently, a rigorous statistical comparison of the technique
                                               to the conventional GC/MS method has been performed.  In this
                                               study, five Missouri soils and one New Jersey soil were analyzed
                                               each  in six replicates by  each  of the techniques. The results,
                                               although not fully evaluated yet, are shown in Table  7.  In this
                                               study, each extract of the MS/MS method was analyzed with no
                                               cleanup, after the  first cleanup step, and after both cleanup steps.
                                               If a  particular run  did  not  meet  internal  quality  assurance
                                                                                              SCREENING TECHNIQUES       57

-------
                           Table 5
      Reproduclbilily of GC/MS/MS System Bused on Repeated
              Analysis of A Standard 10 ppb Solution
Calculated
rile
August
August
August
August
August
August
August
1701
1801
1901
2101
2302
2401
2S01
TCDD
10
10
9
11
11
10
10
, ppb
.05
.26
.75
.42
.00
.72
.69
Percent. Deviation
from Known
0
2
2
14
10
7
6
(10
.5
.6
.5
.2
.0
.2
.9
ppb)







                                                                                    Table 7
                                                            Comparison of MS/MS and GC/MS Method! for TCDD Analysis
WESTON's CC/MS/MS Results,
Soil
Missouri 1
Missouri 2
Missouri 3
Missouri 4
Missouri 5
New Jersey
X
<0.13
1.72
-_-
6.39
16.0
—

A Region VII
IS Results, ppb

-------
        SAFETY AND  HEALTH  INFORMATION  FOR  USE  IN
    RESPONDING  TO  HAZARDOUS  WASTE EMERGENCIES
                                                 JACK ARTHUR
                                          National Library of Medicine
                                               Bethesda, Maryland
INTRODUCTION

  The term "emergency" has been defined by the American Her-
itage Dictionary as a situation or occurrence of a serious nature,
developing suddenly and unexpectedly and demanding immediate
action. Certainly this term  aptly describes those  situations in-
volving leaks, fires, explosions, etc., at chemical waste disposal
sites. The ability of responsible parties  to take the immediate ac-
tion  necessary  for minimizing  damage to life and property de-
pends on their emergency-preparedness.
  Facilitated by passage of CERCLA, Federal and some state gov-
ernments have developed sophisticated  emergency response units,
superbly equipped and organized and  staffed with trained pro-
fessionals.  However, regardless of how well-trained or knowledge-
able emergency staff may be, it is virtually impossible to master the
entire compendium of facts necessary to address the vast array of
potential  concerns  posed by hazardous  chemical  emergencies.
Therefore, the possession of information to support immediate life
and property-saving decisions must be  considered a vital compo-
nent of' 'preparedness''.
  Emergency decisions may be needed in a variety of areas includ-
ing firefighting methods, personal protective equipment and cloth-
ing, evacuation, containment of spills and runoff, inactivation of
reactive substances, ultimate disposal  strategies, etc. Definitive
information in these areas, as  well as  adjunctive, substance-spe-
cific data such as those on chemical/physical properties, reactivity,
flammability/explosivity, human toxicity, ecotoxicity and environ-
mental fate of individual chemicals are essential. Further, each
emergency may involve totally different  sets of chemicals with their
own  set of attendant hazards. Therefore, information to support
the types of decisions above must be available  for a full range of
chemical substances and  compounds. Finally,  the importance of
the decisions and the less-than-ideal circumstances under which
they  must  be made  demand  that information be reliable,  concise
and easily accessible.
  In the remainder of this paper, the author describes some of the
more  important sources  of information for use in responding to
hazardous waste and spill emergencies. Sources are organized ac-
cording to the physical  format in which they  are available, i.e.,
computerized database or hardcopy reference.
COMPUTERIZED DATABASES
  Computerized databases usually can be categorized as either
bibliographic or factual/numeric. Bibliographic databases contain
references  or abstracts of the literature which usually focus on a
given subject area or theme. The National Library of Medicine's
(NLM) Toxicity Information Online (TOXLINE) is a good ex-
ample. TOXLINE contains some 1,500,000 citations, most with
abstracts, covering pharmacological,  biochemical,  physiological
and toxicological effects of drugs and other chemicals.
  Factual/numeric databases contain data from a wide variety of
sources which, through structured selection and formatting, yield
profiles of given subjects, e.g., hazardous chemicals. Because the
salient data have been organized for easy access, this type of data-
base lends itself more to effective utilization in emergency situa-
tions than the bibliographic type. Major factual/numeric databases
relevant to the information needs of emergency response personnel
are described below.

CESARS
  The  Chemical  Evaluation  Search  and  Retrieval   System
(CESARS) contains information on over 180 chemicals; 185 data
fields contain data on physical/chemical properties, uses, produc-
tion volume, acute  toxicity, chronic toxicity,  carcinogenicity,
mutagenicity, teratogenicity, bioaccumulative properties, metabo-
lism,  degradation  products and associated hazards.  Developed
through a joint venture of the Michigan Department  of Natural
Resources and Region V, USEPA, this file contains fully refer-
enced  summarized information from a comprehensive review of the
literature. Available through the USEPA's Chemical Information
System (CIS), CESARS affords the user the ability to manipulate
very detailed information for the purpose of data analyses.

HAZARDLINE

  This database contains information on over 3,200 chemicals,
with data on emergency response, safety, toxicity, signs and symp-
toms,  first aid, regulations and special bulletins for news/current
awareness. Developed by Occupational Health Services, this data-
base is available directly via Tymnet  (Tymshare, Inc.) and STSC
networks. HAZARDLINE data is presented in a handbook format
on a user-friendly, menu-driven system, i.e., the user interacts with
the system at his own terminal to search for, manipulate and  re-
trieve  data specific to his needs.
OHMTADS)

  The Oil and Hazardous Materials Technical Assistance Data
System  (OHMTADS)  contains information on  approximately
1200 substances; 126 fields  contain  data on physical/chemical
properties, biological, toxicological,  cleanup/disposal and com-
mercial data. Maintained by the USEPA and available through
CIS, this database emphasizes information useful in the assessment
of hazards and response activities associated with the release of
harmful substances into the aquatic environment.
                                                                                     SCREENING TECHNIQUES       59

-------
RTECS
  The Registry of Toxic Effects of Chemical Substances (RTECS)
contains information for over 64,000 substances,  with data  on
acute and chronic toxicity,  primary skin and eye irritation, car-
cinogenicity, mutagenicity, teratogenicity, Federal regulations and
status of major tests performed. Maintained by the National  In-
stitute for Occupational Safety and Health (NIOSH), this database
is available via the NLM's Medical Literature and  Retrieval Sys-
tem (MEDLARS) and through CIS.

TDB
  The Toxicology Data Bank (TDB) contains information on over
4,000 substances, with data on toxicity, environmental fate and
effects, safety and handling, physical/chemical properties and reg-
ulations. The TDB has undergone a recent expansion to include
some 140 data elements organized into nine major categories of in-
formation. The expansion was directed  toward the needs of haz-
ardous substance emergency response activities. This comprehen-
sive, peer-reviewed database  is maintained by the NLM and is avail-
able via MEDLARS.

HARDCOPY REFERENCES
  Hardcopy references represent relatively inexpensive, easy-to-use
sources of information. In addition, their portability is highly de-
sirable for use  in the field. However, because of  their physical
form they are much more difficult to update than their computer-
ized counterparts, resulting in long lag times in the addition of new
data. Also, the  static nature  of hardcopy information does not
allow for  rapid collation of data within or among such sources.
Major hardcopy references for use in  hazardous waste emergency
response are described below.
•Catalog Handbook Of Fine Chemicals 1984-1985
  Aldrich Chemical Company
  Published bi-annually by  the Aldrich  Chemical Co., this book
contains information on some 14,000 substances, giving primarily
substance identification and  chemical/physical properties. Indexed
by molecular formula and  chemical name, this book is a high-
quality source for boiling points, melting points and density data.
•Chemical  Hazard  Response  Information  System  (CHRIS)
  Manual 2
  U.S. Department of Transportation, U.S. Coast Guard
  U.S. Government Printing Office, Washington, D.C., 1978.
  Developed by the U.S. Coast Guard, this manual contains  in-
formation  on  approximately 800 substances,  giving chemical/
physical properties,  safety and handling, toxicity, environmental
fate/exposure potential, manufacturing/use and substance identifi-
cation. Intended for use by USCG field personnel, this manual has
become a popular source of information for hazardous waste emer-
gency response.
•Clinical Toxicology Of Commercial Products-4ih ed.
  Gosselin, R.E.,etal.
  Williams and Wilkins Co., Baltimore, MD, 1976.
  This easy-to-use book contains information on 4,288 chemicals,
with data on substance identification, safety and handling, toxicity,
environmental fate/exposure potential and pharmacokinetics.  In-
dexed by  trade name, chemical uses  and manufacturer's name,
this book  is a good source  of information on formulations and
antidote and emergency treatment.
•Dangerous Chemicals Emergency First A id Guide
  Croner Publications Ltd., Surrey, England.
  This guide contains information on signs and symptoms, toxic
hazard ratings and antidote and emergency treatment for 2000
chemicals.
•Dangerous Properties Of Industrial Materials-5lh ed.
  Sax, I.N.
  Van Nostrand Reinhold Co., New York, NY, 1979.
  This book contains data  on toxicity, environmental fate/ex-
posure potential and exposure standards for 15,000 substances. It
is indexed by chemical.
                                                       •Emergency Handling Of Hazardous Materials In Surface Trans-
                                                        portation
                                                        Bureau of Explosives, Association of American Railroads, Wash-
                                                        ington, D.C., 1981.
                                                         This book contains substance identification and response in-
                                                       formation for 2,500 hazardous substances.
                                                       •Environmental Monitoring  Series: Hazardous  Materials  Spill
                                                        Monitoring—Safety Handbook A nd Chemical Hazard Guide
                                                        USEPA, Washington, D.C., 1979.
                                                         This book contains  toxicity, first aid, and exposure protection
                                                       information for 655 chemicals. It was intended for use by spill
                                                       monitors, cleanup personnel and on-scene coordinators.
                                                       •Farm Chemicals Handbook
                                                        Meister Publishing Co., Willoughby, OH, 1983.
                                                         This annually updated book contains substance identification,
                                                       manufacturing/use, toxicity,  chemical/physical  properties  and
                                                       safety and  handling information for 6,500 chemical  and trade-
                                                       name substances. It is indexed by subject but is not referenced.
                                                       •Fire Protection Guide On HazardousMaterials-llh ed.,
                                                        National Fire Protection Association, Boston, MA, 1978.
                                                         This book contains 8,800 flash points, 3,550 mixture reactions,
                                                       1,300 fire hazard properties and 416 hazardous chemical reports
                                                       covering some 10,000 substances. It is referenced but not indexed.
                                                       •Guidelines For The Selection Of Chemical Protective Clothing,
                                                        American  Conference  of  Governmental  Industrial  Hygienists,
                                                        1983.
                                                         This book is a relatively unique compendium of information on
                                                       the impermeability of various protective clothing materials to some
                                                       300 chemical substances. It  is referenced and indexed by chemical
                                                       and chemical class.
                                                       •Hazardous Materials—1984 Emergency Response Guidebook
                                                        U.S. Department of Transportation, Washington, D.C., 1984.
                                                         This book contains useful response data (some generic) for 1400
                                                       substances.  Its coverage of  recommended  evacuation distances is
                                                       notable.
                                                       •Handbook of Chemistry And Physics
                                                        Weast, R.C.,ed.
                                                        CRC Press, Boca Raton, FL, 1979.
                                                         This book is a major source of chemical/physical properties
                                                       for numerous organic and  inorganic chemicals.  It also contains
                                                       information on  analytical methods, is referenced and is indexed by
                                                       subject.
                                                       •Handbook For Environmental Data On Organic Chemicals
                                                        Verschueren, K.
                                                        Van Nostrand  Reinhold Co., New York, NY, 1977.
                                                         This book contains manufacturing/use, chemical/physical prop-
                                                       erties and toxicity data for approximately 2000 organic chemicals.
                                                       It is referenced and indexed by chemical name.
                                                       •Handbook Of Poisoning-\0ih ed.
                                                        Dreisbach, R.H.
                                                        Lange Medical Publications, Los Altos, CA, 1983.
                                                         This book, covering 1200 substances, emphasizes information
                                                       on clinical  findings, signs and symptoms and treatment. It is ref-
                                                       erenced and indexed by subject and chemical.
                                                       •Handbook Of Reactive Chemical Hazards-2nd ed.
                                                        Bretherick, L.
                                                        Butterworth, London, England, 1979.
                                                         Covering over 3000 substances, this book emphasizes informa-
                                                       tion on substance identification and hazardous reactions. It is ref-
                                                       erenced and indexed by chemical and subject.
                                                       •Handbook of Toxic and Hazardous Chemicals
                                                        Sittig, M.
                                                        Noyes Publications, Park Ridge, NJ, 1981.
                                                         Covering some 600 substances, this book contains concise chem-
                                                       ical property data and health and safety information useful to pro-
                                                       fessionals who must make expeditious public health decisions.
                                                       •Herbicide Handbook-4th ed.
                                                        Weed Science Society of America, Champaign, IL, 1979.
60
SCREENING TECHNIQUES

-------
  This book contains data on environmental fate and antidote and
emergency treatment for exposure to 144 substances. It is refer-
enced and indexed by chemical and product names.
•Herbicides: Chemistry, Degradation, And Mode OfAction-
 2nd ed.
 Kearny, p.c. and Kaufman, D.D., eds.
 Marcel Dekker, Inc., New York, NY, 1975.
  This book  emphasizes information on environmental fate for
 more than 1000 herbicides. It is referenced and indexed by author,
chemical and subject.
•Material Safety Data Sheets-General Electric Co.
 Nielsen, J.M.,ed.
 General Electric Co., Schenectady, NY, 1980-83.
  This compendium of material safety data sheets (MSDA) con-
tains a full range of information on 523  chemicals and 275 trade-
name substances with about 100 new chemicals added per year.
Each MSDS is referenced and the compendium is indexed by chem-
ical.
•Merck Index-lOth ed.
 Windholz, M., et al. eds.
 Merck and Co., Inc., Rahway, NY, 1983.
  This book contains data on 10,000 chemicals and covering sub-
stance identification, manufacturing/use, safety and handling and
toxicity. Its coverage of chemical/physical properties is notable. It
is referenced and indexed by CAS name and number.
•NIOSH/OSHA Occupational Health  Guidelines For Chemical
 Hazards
 Mackison, F.W., etal. eds.
 U.S. Government Printing Office, Washington, D.C., 1981.
  This guideline series contains data on 398 chemicals covering
substance identification,  chemical/physical properties, toxicity and
safety and handling. It is referenced but not indexed.
•POISINDEX
 Rumack, B.H.,ed.
 Micromedex, Inc., Denver, CO.
  Available only in microfiche, this source nevertheless contains
extensive information on the treatment of approximately 575 toxic
substances. It is updated quarterly.
DISCUSSION
  Noteworthy advances have been, and are continuing to be, made
in compiling, organizing and delivering hazardous waste and spill
emergency response data. However, certain undeniable weaknesses
still exist. In general, even the major sources focus primarily on
the high production-volume chemicals, leaving extensive gaps in a
responder's ability to deal with the lesser characterized substances
which also find their way into disposal sites.
  Further, information on mixtures,  including popular trademark
compounds, is grossly insufficient. Also, no single source contains
the complete menu of  necessary data. As a  consequence, emer-
gency response personnel often must  rely on a number of sources,
thus diminishing available response time.  Chemical-specific data on
reactivities,  inactivation  procedures, detailed  cleanup/disposal
methods  and suitable (impermeable)  material for protective cloth-
ing are all in need of further development. Lastly, because of logis-
tical and  economic difficulties, most  sources are, at best, only in-
frequently updated.
  The foregoing problems are not insurmountable. Continual ad-
vances in computerized delivery systems, research to fill data gaps
and Federal support are all contributing to  improvements. In addi-
tion, the  Occupational  Safety and Health's (OSHA) new Hazard
Communication Standard (29 CFR 1910.1200) should establish a
standardized, comprehensive base of chemical information.  The
standard  covers all chemicals produced, imported or used  within
the United States' manufacturing sector and requires that employ-
ers  provide  labeling, material safety data sheets and employee
training.
                                                                                        SCREENING TECHNIQUES
                                                          61

-------
     METHODOLOGY FOR SCREENING  AND EVALUATION
                          OF  REMEDIAL  TECHNOLOGIES

                                             VIRGINIA HODGE
                                           KATHLEEN WAGNER
                                           PAUL ROGOSHEWSKI
                                                JRB Associates
                                               McLean,  Virginia
                                            DOUGLAS AMMON
                                    Solid and Hazardous  Waste Division
                                   U.S. Environmental Protection Agency
                                               Cincinnati,  Ohio
INTRODUCTION

  Sections 300.68(g), (h) and (i) of the National Contingency Plan
(NCP) outline a three-level process  for selection of the most ap-
propriate remedial actions for a given site. First, a limited number
of remedial  action alternatives are developed based on site prob-
lems. Second, an initial screening of each alternative is made to
eliminate those which are clearly  inappropriate or infeasible. The
third level of the process involves a detailed analysis of a limited
number of alternatives which remain after the initial screening. One
goal of the NCP is to ensure, to the extent practicable, that these
remedial alternatives meet the need for protection of public health
and welfare and the environment in the most cost-effective manner.
Therefore, these broad criteria have been established for analysis
during the remedial action selection process: acceptable engineering
practices, costs; and public health, environmental and institutional
effects.
  In response to these requirements, a methodology was developed
to provide  guidance  for  the  process  of  screening  remedial
technologies and developing  remedial action alternatives. This
methodology was prepared  as  a  guidance  document  entitled
"Methodology  for  Screening and  Development  of Remedial
Responses"  and is summarized in this paper.
  An overview of the elements of the methodology, consisting of
three steps, is shown in Figure 1.  These important steps are:
•Identify general response actions
•Technology screening and alternatives development
•Technical evaluation of alternatives
                                                      The evaluation of remedial action technologies and alternatives is
                                                    a highly complex  process and in many instances relies on best
                                                    engineering judgments. As a result, it is not the intention of this
                                                    methodology to recreate in detail the thought process used in the
                                                    screening and evaluation of remedial technologies and alternatives.
                                                    Rather, the intent is to provide a system that can track and prompt
                                                    the decision process involved in the technical screening of remedial
                                                    action technologies, the development of remedial alternatives and
                                                    the technical ranking of remedial alternatives.
                                                      To be more effective, the methodology should be integrated with
                                                    the  guidance  forthcoming from  the USEPA  for  conducting
                                                    remedial investigations  and feasibility studies under CERCLA.

                                                    IDENTIFY GENERAL RESPONSE ACTIONS

                                                      The First step is to identify existing site problems and the general
                                                    response actions that may be applicable to remedying site prob-
                                                    lems. A general response action represents a group of remedial
                                                    response technologies (i.e., air pollution controls, direct treatment)
                                                    relevant to a specific site problem.
                                                      Definition of site problems relies on existing data collected in
                                                    preliminary site studies for Superfund site ranking and evaluation
                                                    efforts. The site problems are then matched to general response ac-
                                                    tions. A matrix for identifying applicable response actions based on
                                                    site problems is given in  Figure 2. There are  10 response action
                                                    categories covering potential response elements.  The process of
                                                    matching site problems with response actions is an obvious first
                                                    step and is critical to the overall remedial action selection process
                                                      Figure 1
                           Technical Screening and Evaluation of Remedial Technologies and Alternatives
62
SCREENING TECHNIQUES

-------
because ii eliminates inapplicable responses, thereby limiting the
universe of potential response actions and simplifying and focusing
turther screening and evaluation.

TECHNOLOGY SCREENING
  The second step is to identify and screen potentially applicable
technologies  and then develop remedial alternatives. To do this,
one first identifies feasible technologies to deal with specific prob-
lems. The methodology manual provides  a comprehensive list of
over  150  remedial  technologies classified  according  to general
response action categories. The general response actions thus deter-
mine "families" of potentially applicable technologies.






Site Problem
Volatilization of chemicals into air
Hazardous particulates released to
atmosphere
Dust generation by heavy construction
or other site activities
Contaminated site run-off
Erosion of surface due to wind or water
Surface seepage of leachate
Flood haiard or contact of surface
water body with wastes
Leachate migrating vertically or
horizontally
High water table which may result in
gruundwatar contamination or inter-
fare with other remedial technologies
Precipitation infiltrating into site to
form leachate
Evidence of methane or toxic gases
migrating laterally underground
On- sue waste materials in non-
disposed form- drums, lagooned
waste, wastepiles
Contaminated surface water, ground-
water or other aqueous, or liquid
waste
Contaminated soils
Toxic and/or hazardous gases which
have been collected
Contaminated stream banks and
sediments
Drinking water distributor system
contamination
Contaminated wwer lines



1
u
c
.2
1

•
.

.






























o
Ł
3
5
s
8
r
3
(A





•
•
•
.






.


















i
o .
12
• c
• o
if"
If










.

.


.


















o
0
o
c
.2
5
a
i
0

















.
















3 •
uj o
(/) (C
C C
• c
Ii



















.





•







S «
5 '5
If
I'
si
« o
Ii
ucc




























•






c
I
z
3
o5
Ł










.








.


.


*








C
1
1
»
*
5



















.





•
.

•





B
1 =
:»
o ,.
J;
Q 5
e E



















.


.


•


•




\
• ±
1 2
fi
*.&
\\
r.
n






























•

•
                            Figure 2
     Matrix of General Response Actions for Specific Site Problems

   In the screening process, one uses a series of screening tables to
eliminate inappropriate technologies;  this judgment is based solely
on technical factors. A separate series of screening tables is pro-
vided for each of the ten general response action categories iden-
tified. Within each table, all potentially applicable technologies in a
particular response category are listed and briefly described. These
technologies are then  further broken down into technology op-
tions.
   An example screening table for a partial technology, listing of the
general response action category, surface water controls, is shown
in Table 1. One technology is capping. Technology options within
this category include clay cap, asphalt cap and synthetic liner.
   Each screening table contains three  additional columns that pro-
vide necessary information to screen technologies and subsequently
develop alternatives.  Under the  second  column, in Table  1,
"Technical Factors Precluding Implementation," there is a list of
limiting site, waste and technology characteristics which preclude
implementation of the technologies. For example, the presence of
very low permeability soils would be a site limitation that would ex-
clude the use of in situ treatment methods since it would be difficult
to ensure complete mixing of treatment reagents with waste com-
ponents. Also, the presence of a strongly acidic groundwater plume
would  be a waste  limitation  that  would exclude the  use  of a
downgradient, soil-bentonite slurry wall since  the wall would be
degraded by contact with acids.  Further, technology limitations
might include inherent operation, construction and  maintenance
problems. An example of a technology limitation would be the in-
ability to dredge below 65 ft with a hydraulic dredge since that is the
maximum reach  for that type of equipment.
  In addition to listing  those factors precluding implementation,
the screening tables also list "Critical Factors Affecting Selection."
These factors do not necessarily preclude implementation under all
circumstances, but they are listed in order to identify certain factors
that  need to be considered during screening  and/or during the
detailed evaluation of alternatives. Some of the factors listed under
this column may affect performance or implementability to such an
extent that a technology may be eliminated during technical screen-
ing.  On the other hand, these factors may raise a "red  flag" for
concerns that need to be carefully considered during the additional
screening of alternatives (cost, public health and environment). Ex-
amples of factors affecting selection are as follows:
•Design  and operational considerations that  significantly affect
 implementation and  performance of technologies. For  example,
 asphalt caps are subject to cracking. This limitation may make it
 difficult to maintain and monitor performance  of an asphalt
 cap, particularly at a very large or remote site, and may be a basis
 for eliminating  this technology option.
•Technology status is an item which requires close consideration
 during the screening process since the National Contingency  Plan
 requires that technologies used for remedial actions be developed
 and demonstrated. However, in certain situations,  technologies
 under development may be considered for application.
•The cost and impacts of associated technologies may reduce the
 feasibility of a particular technology. For example,  the leaching
 of metals from a fly ash cap may preclude the  implementation of
 this technique where the impacts would be unacceptable. Also,
 the requirements for specialized equipment to implement  certain
 solidification/stabilization technologies may  make these tech-
 nologies cost prohibitive  or unavailable under many circum-
 stances. These are factors which will need to  be considered fur-
 ther in the cost and impacts screening steps.
  The final column in the screening table is designated "Additional
Technologies" and includes a listing of both associated and secon-
dary technologies. An associated technology is a technology that
may be used together with the technology being screened to im-
prove its performance. For example, a cap and a pumping system
or subsurface drainage system are frequently used together with a
circumferential slurry wall to minimize infiltration and prevent the
bathtub effect or overtopping of a cap. A secondary technology is
one  that would be required to handle a secondary contamination
problem that occurs  as a direct result of  implementing the
technology  being screened.  For  example, implementation  of a
groundwater pumping system generates an aqueous waste stream
which  generally  requires treatment.  Therefore,  aqueous waste
treatment would be considered a secondary technology. Combining
associated and secondary technologies with the primary technology
for resolving a site problem is actually the first step in building
alternatives.
  Technologies that have passed the technology screening process
can then be combined to form overall alternatives that address all
site  problems identified. As mentioned previously, the  screening
tables provide a useful  starting point in formulating alternatives
from primary, secondary and associated technologies. This  may be
the only step required when  a site has only one problem. More
often,   multiple  site problems  exist, so  diverse,  compatible
technologies must be combined to address all site problems. In
either case, a workable number of remedial alternatives must be
developed based on feasible technologies that appear to adequately
address all site problems.
                                                                                           SCREENING TECHNIQUES
                                                           63

-------
                                                               Table 1
                                       Example Screening Table for Surface Water Controls (Excerpt)
Technology

100. Capping
    Purpose: To
    control surface
    and prevent
    water infiltration
               Technical Factors Precluding
               Implementation
               1. Not applicable to areas with
                 very steep slopes (  25%)
                                                      Factors Affecting
                                                      Selection
1. All capping materials are subject to
  degradation through ground subsidence
2. Lagoons must be Tilled and/or regraded
  prior to capping
3. Require regular inspection for burrowing
  animals and growth of deep-rooted plants
Additional
Technologies

Associated
I. Grading (300) to control run-on and run-off
2. Revegetation (400) to prevent erosion

3. Gas collection systems (1000) to reduce build-
  up of hazardous or toxic vapors
4. Subsurface drains (800) to control leachale and
  prevent overtopping of cap
Secondary
1. Gaseous treatment (1600) of collected gases
2. Aqueous waste treatment (1700) of collected
leachale
 101. Clay
 102. Asphalt
 103. Synthetic liners
               I. Not favorable in arid climates
               2. Not suitable for direct contact
                 with organic or inorganic acids
                 and bases

               1. Not favorable in areas where
                 exposure to heal is excessive,
                 such as arid climates
               2. Not suitable for direct contact
                 with high-concentration
                 organics
               1. Not favorable in areas where
                 exposure to heat is excessive
               2. Not suitable when in contact
                 with high concentrations of
                 incompatible  organics
                                                      I.  Cracking (repairable)
                                                      2.  Chemical and photodegradation
                                                      3.  Contact with metals and soluble organics
                                                         may be a problem
1. May require extensive subgrade preparation
2. May require special cover
3. Subject to tearing and degradation through
  sunlight and exposure
                                          Associated
                                          I.  Covering with another soil layer prior to re-
                                             vegetation to maintain moisture in cap and
                                             prevent deterioration due to penetration by
                                             vegetation
                                          Associated
                                          I.  Polypropylene fabric underlines (103)
Associated
1. Covering with another soil layer to protect
  from degradation through exposure
   Given that  a large  number  of technologies  may pass  the
 technology screening process and that a site may have a number of
 primary and secondary contamination problems, the number of ap-
 plicable technologies and feasible alternatives could be very large.
 Then, best engineering judgment must be used to limit the number
 of  alternatives  considered  for  detailed  evaluation.  The
 methodology  manual describes  several general  approaches  for
 limiting the number of alternatives without affecting the credibility
 of the feasibility study. These methods  involve various techniques
 to combine the permutations that may exist between and among
 technologies.
   Once  the  remedial  technologies have  been  screened  and
 developed into remedial  alternatives, these alternatives undergo an
 initial screening based on cost,  health and environmental  criteria.
 This screening is conducted under the feasibility study and serves to
 limit the  number  of remedial  alternatives  undergoing  detailed
 analysis by eliminating alternatives that  do not sufficiently address
 cleanup requirements. The alternatives remaining after this screen-
 ing  undergo the final, detailed evaluations.

 TECHNICAL EVALUATION  OF ALTERNATIVES
   The third and final step is to determine the technical suitability of
each alternative  for dealing with the site-specific problems. Each
 alternative from the initial screening process is evaluated and then
 rated relative to the other alternatives with respect to performance,
 reliability, implementability, time and safety. As shown below, two
 measures have been designated  for determining each of these five
 criteria:
 •Performance
  -Effectiveness
  -Useful life
 •Reliability
  -Operation and maintenance requirements
  -Demonstrated and expected reliability
 •Implementability
  -Site conditions affecting constructability
  -External factors affecting implementation
                                                             •Time
                                                              -Time to implement
                                                              -Time to achieve beneficial results
                                                             •Safety
                                                              -Safety and health of workers
                                                              -Safety and health of nearby communities
                                                               Each of these measures is described in the guidance document,
                                                             and a list of questions intended to assist in the evaluation of each
                                                             alternative is provided. For  example, questions that can be used to
                                                             evaluate the effectiveness of a given alternative include:
                                                             •How effective is  the technology/alternative in meeting  site ob-
                                                              jectives (e.g., volume of contaminated material removed or con-
                                                              tained; level of treatment  achievable; volume of water diverted
                                                              or collected)?
                                                             •Are there any site or waste characteristics which could potentially
                                                              impede effectiveness?
                                                             •Is there any particular technology within the alternative which
                                                              is limiting in terms of effectiveness?
                                                             The technical evaluation should include a written response for each
                                                             of these questions, where applicable, to support the analysis pro-
                                                             cess.
                                                               Once  a written  response has been prepared for each  of the
                                                             preceding questions,  the technical ranking of alternatives can be
                                                             performed using Figure 3. The alternatives can be ranked in order
                                                             of their relative desirability with  respect to each criterion. Alter-
                                                             natively,  each  remedial  alternative and technology  can be rated
                                                             with respect to  the  absolute  degree  to which the  alternative or
                                                             technology effectively fulfills each criterion. If the relative evalua-
                                                             tion method is chosen, the highest number is generally the number
                                                             of alternatives  under consideration.  If the  absolute evaluation
                                                             method is chosen,  the numerical values are generally ranges (such
                                                             as 1  to  10; 1 to 5; or -1,0, +1) with the lowest number represen-
                                                             ting a "base-line" alternative. Whichever is chosen, there should be
                                                             a consistent numerical ranking with the highest number indicating
                                                             the most desirable alternative or technology under each criterion.
                                                               One possible  exception to the above scoring is the criterion for
                                                             time. If so desired,  the numerical value  for  time  could  be the
64
SCREENING TECHNIQUES

-------

                           Figure 3
              Technical Feasibility Ranking Summary
ACKNOWLEDGEMENTS
  This paper was developed under the sponsorship of the USEPA
Solid and Hazardous Waste Research Division, Cincinnati, OH.
USEPA contract #68-03-3113, work assignment #8-2.
number of months or years relevant to each technology or alter-
native. The overall time could be the sum of the time required for
implementation and achieving beneficial results. Note that the time
to achieve beneficial results should not  include implementation
time.
  To further support  the  ranking  of  alternatives, Figure  4,
"Technical Evaluation Summary Sheet," is used to briefly sum-
marize the major strengths and weaknesses of each  alternative in
terms of the evaluation criteria. This summary provides the public
with a means of understanding the rationale used in ranking alter-
natives.
                                                                           Alternative Name:
                                                                           Alternative Description:
                                                                           Performance
                                                                             Effectiveness:
                                                                             Useful Life:
                                                                           Reliability
                                                                             Operation and Maintenance Requirements:
                                                                             Demonstrated and Expected Reliability:
                                                                           Implementability
                                                                             Site Conditions Affecting Constructability:
                                                                             External Factors Affecting Implementation:

                                                                           Time
                                                                             Time to Implement:
                                                                             Time to See Beneficial Results:
                                                                           Safety
                                                                             Safety and Health of Workers:
                                                                             Safety and Health of Nearby Communities:
                                                                                             Figure 4
                                                                                 Technical Evaluation Summary Sheet
  Concurrent with the technology evaluation are the analyses of
the alternatives regarding cost,  public health and environment.
These four analyses  will be combined in the feasibility study to
develop the information to select the cost-effective remedial alter-
native.
CONCLUSIONS
   The  methodology  presented  in  the  procedures  manual,
"Methodology  for  Screening   and  Evaluation  of  Remedial
Responses," is a three-step analysis method to be used in remedial
investigations  and feasibility studies conducted in the Superfund
program. The methodology is a system for documenting the iden-
tification,  screening,  elimination  and  selection  of  remedial
technologies and remedial alternatives. This method aids in  the
decision-making  process  by  providing a  formal,   analytical
framework that can be recreated for an individual site and provide
consistency in approaching different sites.  Finally,  the selection
methodology  serves to generally structure a process that, in  the
past, has been more intuitive or conceptual in nature.
                                                                                           SCREENING TECHNIQUES       65

-------
A SUPERFUND SITE  ATMOSPHERIC STUDY: APPLICATION
             TO REMEDIAL RESPONSE  DECISION-MAKING

                                                 T. IACCARINO
                                                   R. SIGNER
                                                   R. JUBACH
                                                   D. SMILEY
                                                NUS Corporation
                                             Gaithersburg, Maryland
 INTRODUCTION

  Uncontrolled hazardous waste disposal sites vary considerably in
 the type, size, nature and amount of hazardous substances con-
 tained and in the nature and severity of hazards presented. They in-
 clude landfills containing loose or containerized wastes and open
 dumps of barrels and drums. The hazardous substances may in-
 clude organic solvents,  waste oils, pesticides,  heavy metals, in-
 organic acids, alkalis, salts, explosives, flammables, radioactives,
 carcinogens and infectious materials.


 DESCRIPTION OF PROGRAM

  The Lackawanna Refuse Site is located west of Keyser Avenue,
 Old Forge, Lackawanna County, Pennsylvania  (Fig. 1). The site
 lies on the border between the Borough of Old Forge and Ransom
 Township.
  This 258-acre Superfund site is located in a hilly area previously
 deep-mined and strip-mined  for coal. The  site area is hilly and
 forested except for an open area near the center of the site where
 the main disposal activity occurred. Open strip cuts were used for
 municipal and commercial waste disposal; industrial and hazardous
 wastes were also dumped in the same location. In addition, liquid
 wastes were allegedly dumped along the access road and in a strip
 cut along the access road known as the borehole pit.
  The site is inactive.  Surrounding land use includes former strip
 mining for coal,  rural and agricultural activities and some residen-
 tial use.  Hunting  occurs in  areas around  the site. The site is
 bordered by a few houses to the  east and  a trailer park to the
 southeast. The Austin Heights section of Old Forge Borough is
 northeast of the site. The area west of the site is hilly and forested.
 There is a population of approximately 9,500 within a I -mile radius
 of the site, residing in the Borough of Old Forge.
  As a result of the USEPA decision to pursue remedial action at
 the Lackawanna Refuse Site, a thorough remedial investigation was
 conducted. The purpose of the investigation was to assess the threat
 to public health, welfare and the environment posed by the site and
 to identify potential options to remedy the problem. As part of the
remedial investigation, a  feasibility study is currently being con-
ducted; this study involves a  detailed evaluation of feasible alter-
natives using economic, environmental and engineering factors.

 Contamination Potential

  A consideration in the feasibility study is the various dispersal
 pathways by which hazardous waste could  potentially affect the
 surrounding area. Most  common pathways  are  through the soil,
                                                     groundwater, surface water,  direct contact and  the atmosphere.
                                                     The Lackawanna Refuse Site is unique because the atmospheric
                                                     pathway, in conjunction with the surrounding complex terrain,
                                                     presented a potentially signficant threat to the nearby surrounding
                                                     population when remedial action alternatives were considered.
                                                       In addition, an assessment of the air pathway became important
                                                     for the preparation of emergency preparedness procedures for use
                                                     at the site during planned exploratory excavations. Development of
                                                     these procedures was necessary in the event of an accidental release
                                                     of toxic substances to the atmosphere (e.g., breaking of a barrel of
                                                     a liquid  toxic  chemical  and subsequent volatilization  of  the
                                                     substance).
                                                     Possible Air Pollution

                                                       Therefore, to assess the potential impact along the air pathway,
                                                     an atmospheric field study was performed at the site. The objective
                                                     of the field study was to characterize atmospheric transport and
                                                     diffusion conditions in the near-vicinity of the site associated with
                                                     potential ground-level releases of pollutants.
                                                       Local wind patterns may be complex at Lackawanna due to the
                                                     influence of terrain features in and near the site. Thus, standard
                                                     methodologies for evaluating atmospheric transport and dispersion
                                                     may  not be applicable. The objective of this field study is to
                                                     characterize atmospheric transport and diffusion conditions in the
                                                     near-vicinity of  the  site  associated  with potential ground-level
                                                     releases. Specifically, the goal  was  to characterize airflow and
                                                     dispersion  during those meteorological conditions that have  the
                                                     greatest  potential  for impact on nearby neighborhoods. Those
                                                     meteorological conditions can be classified into three categories:
                                                     •Regional flows from the southwest quadrant (the prevailing wind
                                                      conditions at the site)
                                                     •Regional  flows  from the northwest  quadrant (a secondary flow
                                                      condition)
                                                     •Local flows from the  west  quadrant (generally associated with
                                                      drainage downslope flows during periods of low regional wind
                                                      speeds)
                                                       Regional winds from  the southwest occur during all months of
                                                     the year and are the predominant wind flow direction for every
                                                     month except March.  Regional  northwest winds  are common
                                                     during the late fall, winter and early spring months and are the
                                                     prevailing  wind direction during the month of March. Drainage
                                                     flow  conditions, when  they do occur, can be expected to occur
                                                     predominantly at night  and are most prevalent during the spring
                                                     and fall and, to some degree, during the summer. The frequency is
                                                     lower in the winter months. Daytime  drainage flow conditions are
66
AIR MONITORING

-------
          -„.-_,_  -.-. o
                                                            Figure 1
                                                      Lackawanna Site Area
rare, although possible, during the winter months if a snow cover is
present, or near sunrise or sunset during any season.
  The Dec. 6 through 20, 1983, atmospheric field program at the
Lackawanna site consisted of three components:
•Meteorological monitoring
•Smoke releases
•Tracer releases
  The technical approach for the tracer program involved release
of sulfur hexafluoride (SFg) as a tracer gas at a known release rate.
A network of SF6 samplers was deployed at locations downwind of
the  release. Concentrations measured at these sampling locations
can be related to the source in order to determine relative dilution.
These values  can  also  be evaluated  with wind  and  stability
measurements to characterize atmospheric transport and dispersion
conditions. Smoke releases were also used to visually observe the
plume and provide qualitative transport information. Results from
this field study provide a basis for estimating potential concentra-
tions at  locations  in the site vicinity in the event of an accidental
release.
  Another paper is planned to address the technical aspects of the
tracer study.
CONCLUSIONS
  An atmospheric study has been completed for the Lackawanna
Refuse Site. The analyses indicate that off-site meteorological data
are not representative of the Lackawanna area atmospheric condi-
tions due to the complex terrain at the site. Results from the study,
however, provide a basis for characterizing atmospheric dispersion
and transport in the site area. This study has facilitated the use of
standard dispersion models with appropriately modified input and
output data.  The atmospheric study and  subsequent modeling
results have been applied to remedial action decision-making. This
process included evaluation of air-pathway impacts associated with
alternative site cleanup options. Another application has been for
emergency preparedness associated with investigative excavations.
                                                                                                  AIR MONITORING
                                                                                                                            67

-------
                   THE EFFECT OF  WIND  SPEED  ON  THE
      EMISSION  RATES OF  VOLATILE CHEMICALS FROM
                 OPEN  HAZARDOUS WASTE DUMP SITES

                                           JACK CARAVANOS
                                         Hunter College of CUNY
                                           New York, New York
                                        THOMAS T. SHEN, Ph.D.
                              Columbia  University School of Public Health
                                           New York, New York
INTRODUCTION

  The increased recognition of widespread mismanaged hazardous
waste disposal sites in the United States has caused environmental
and public health  officials to seek  ways  to  identify ambient
chemical exposures and evaluate their public health importance. In
the past, the environmental health impact from these sites focused
primarily on the  contamination of local water supply  aquifers.
Recently, however, attention has been directed toward hazardous
air emissions from uncontrolled  hazardous waste sites.1'2 The slow
emission rates  characteristic of this  type of environmental con-
tamination may produce long term chemical exposures  which may,
in turn, affect the health of neighboring communities.
  Theoretical mathematical models have been developed to predict
such emissions.3-4 One such model' attempts to describe the surface
evaporation of chemicals from saturated soil. Evaporation and dif-
fusion under these conditions is a strong function of  wind speed
and temperature.
  In this paper, the authors present  the results of laboratory ex-
periments conducted to evaluate the  effect of wind speed on the
chemical emission rates from different soil types. With  proper con-
sideration of wind speed, this model  may be used in assessing the
air emissions  and, therefore, the public health impact of open
chemical dumps or land-based spills.
BACKGROUND

  In 1855, Adolph Pick* introduced a theory which stated that the
diffusion of chemicals within various  medias was similar to  heat
and electrical conduction:
                    dx
       Jz = - cD
                    dz
                                                    (1)
Where:
        Jz = Flux Rate per Area (cm2)
         c = Molar Density of the Fluid (g)
        D = Diffusion constant (cmVsec)
         z = Distance (cm)
         x = Mole Fraction of Chemical in Fluid or Soil (g)
  A critical element of this equation is the diffusion constant, D.
Several researchers have developed methods for calculating dif-
fusivities from chemical data'. The method used here is Gilliland's
et al. 's modification and is described by Shen.3'6
                                                                    0.001T L
                                                                         M^frv,)*]'
                                                                                                              (2)
                                                          Where:
            T =
                Temperature (degrees K)
      Ml, M2  = Molecular weights of components 1 and 2
                 (g/mole)
       VI, V2  = Atomic Diffusion Volumes of components 1
                 and 2 (cmVmole)
           P  = Chemical Vapor Pressure (mm Hg)
  Since diffusivity is a strong function of temperature, emission
and evaporation rates  will vary considerably under  varying  en-
vironmental conditions. Shen' lists a number of the chemical dif-
fusivities  for many environmental contaminants at different
temperatures. These are  useful when calculating  theoretical air
emissions from hazardous waste sites.
  Ziegler' modified Fick's Law for  application to gaseous emis-
sions from surface chemical spills in the form  of open pools of liq-
uid. Arnold1 included  a wind speed component to the equation
which Shen' modified to consider land based spills. The form of the
equation used in this paper is:
                                                            Emission Rate = 1 cw
                                                              (cmVsec)
                                                                                              m
                                                                                                              (3)
Where: c = Chemical Vapor Pressure (atm)
      w = Width of Land Spill or Landfill (cm)
      D = Diffusivity (cmVsec)
      L = Length of Land Spill of Landfill (cm)
      V = Wind Speed (cm/sec)
      F = Vapor Pressure Correction Factor
      m = Weight of Chemical in Soil (g)
      M = Weight of Chemical and Soil (g)


  The above equation predicts the quantity of chemical expected to
evaporate under given wind speed and temperature conditions. For
determining the correction factor (F) see Figure 1. The Equation
below can be used to convert the emission rate into mass units:
                                                            Emission Rate
                                                                (g/sec)
                Emission Rate x
                      (cmVsec)
                                MW
                                MV
(4)
                                                          Where:  MW = Molecular Weight (g/mole)
                                                                  MV = Molecular Volume (cmVmole)
68
        AIR MONITORING

-------
     the wind speed variable (V) of Equation 3 is isolated and ex-
pressed in exponential form, it can be seen that emission rate has
oeen related to the wind speedvto the 0.5 power:

  Emission Rate = 2cw ™.
                       M
                                                           (5)

or
  Emission Rate = K * VX
  If the chemical and soil terms are defined as the constant K and
Equation 5 is rearranged, then exponential wind speed term X can
be expressed as:
Where:
        x = ln(E) -  ln(K)
                   ln(V)

        X  = Exponential Wind Speed Constant
        E = Emission Rate (cmVsec)
        K = 1  Chemical and Soil Constant
                                                           (6)
                20      40      60      BO
                  VAPOR PRESSURE (percent)

                           Figure 1
                   To Find Correction Factor F
                                                100
Date Source: Reference 8
  By experimentally determining the actual value of the wind speed
exponent, Equation 3 becomes more valuable in predicting the
chemical emission rates as wind speed varies.  However, the rates
predicted  using Equation 3  are only applicable to  the  surface
evaporation of chemicals from soils. Once  the  top layer of
chemically contaminated soil evaporates, other equations  may be
used to predict sub-surface emission.4-9

METHODS AND MATERIALS

  In order to calculate the wind speed function, experiments were
set up to collect the necessary data. The experiments involved deter-
mining the actual emission rates for three commonly used  organic
solvents under varying conditions. The chemicals, benzene, carbon
tetrachloride and trichloroethylene, are commonly  used solvents
and  have  been prevalent contaminants in hazardous  waste sites.
Three different soils were evaluated under three different wind con-
ditions. Each experiment was done in triplicate to assure accuracy
and reproducibility. The variables were:
                    -Clay, Sand and Organic Topsoil
                    -0.5, 2.5 and 5.0 miles/hr (0.8, 2.4 and
                        8.0 km/hr)
                    -Benzene, Carbon Tetrachloride  and
                        Trichloroethylene
   Wind Speed

   Chemicals
  Some chemical properties of the test compounds are shown in
Table 1.

                           Table 1
             Chemical Properties of Test Compounds'
Properties
Formula
Mol. Weight
Boiling Pt. °C
Vapor Pressure
mm Hg at 20 °C
Density at 25 °C (g/ml)
Dif fusivity at 25 °C
(cm2/sec)
Benzene
C6H66
78.11
80.0

74.5
0.87

0.088
Carbon Tetra-
chloride
CC14
153.82
76.9

87.4
1.58

0.082
Trichloro-
ethylene
CHC1:CC12
131.39
87.1

60.0
1.46

0.084
                                                                    The experiments involved saturating a known amount of soil in a
                                                                  shallow stainless steel evaporation pan and measuring the loss  of
                                                                  chemical using gravimetric procedures. The soils were pre-dried  so
                                                                  that any change of weight would be attributed to vaporization and
                                                                  vapor diffusion leaving the tray and soil. The loss was measured  at
                                                                  precise 15 min. intervals and this, together with other chemical, soil
                                                                  and wind speed information, was used to calculate the wind speed
                                                                  exponent.
                                                                        IT
                                                                         i
                                                                         |
                                                                         ii
                                                                         ^
                                                                         r
                                                                         jf
                                                                         1L
                                                                                            Figure 2
                                                                             Plot of Particle Size Versus Percent Retained
                                                                    The three soils tested varied significantly with respect to density,
                                                                  porosity and particle size distribution (Table 2). The sandy soil was
                                                                  the least porous and the organic topsoil the most. The topsoil was
                                                                  also clearly more organic than the other two and had a wider range
                                                                  of particle distribution (Figure 2).
                                                                                             Table 2
                                                                                 Physical Characteristics of Test Soils
                                                                       Soil Type
                            Porosity (%)
Density (g/ml)
Clay
Sand
Topsoil
48
32
51
1.34
1.59
0.96
  Each experiment was repeated under three varying wind condi-
tions. These were simulated using various methods and proved con-
sistent and accurate throughout the experiment.  Verification  of
wind speed was done using a calibrated Alnor Velometer.
  Measurements were made repeatedly at the air-soil interface dur-
ing each experiment with little variation observed. Evaporation
rates  were determined by saturating a known  volume  of pre-
weighed moisture-free soil (750 ml) and measuring the weight of
                                                                                                  AIR MONITORING
                                                                                                                            69

-------
                       S-
                         -S
                          r -
                       8 J
                                             234

                                                  TIME   (hours)
                                                             Figure 3
                                   Evaporation of Benzene from Clay at 0.5, 2.5 and 5.0 mph Wind Speed
 change at specific intervals. The evaporation trays used were 1.45
 cm x 1.95 cm x 0.31 cm with a surface area of 282 cm3. Evaporation
 rates vary with temperature, therefore measurements were taken at
 the beginning and at various intervals throughout each experimen-
 tal run.

 RESULTS

   The initial, or first 15 min, emission rates (g/min) for  each  ex-
 periment are listed in Table 3. Figure 3 contains a plot of the emis-
 sion rate for benzene in clay at all three wind speeds. The shapes of
 the emission curves for the other combinations of soil and chemical
 are similar to Figure 3.
                            Table 3
 Observed Surface Emission Rales of Benzene, Carbon Tetrachloride and
 Tricbloroethylene in Clay, Sand and Topsoll During (he First 15 Minute
             Evaporation Interval in Grams per Minute

                                          Emission Rales
Chemical
Benzene


Carbon
Tetrachloride

Trichloro -
ethylene

Wind
(mph)
0.5
2.5
5.0
0.5
2.5
5.0
0.5
2.5
5.0
Speed
(km/hr)
0.8
2.4
8.0
0.8
2.4
8.0
0.8
2.4
8.0
Clay
0.867
2.200
4.533
1.467
4.200
9.533
0.933
2.733
6.000
Sand
0.933
2.200
3.667
1.667
4.000
7.333
0.800
2.600
5.333
Topsoil
0.667
2.333
4.600
1.400
4.800
11.133
0.800
3.133
7.933
  The observed emission rates of Table 3 (after being converted to
cmVsec) were used in Equation 6 for the calculation of the ex-
ponential wind speed function. In calculating the wind speed cons-
tant of  Equation 6, the diffusivity, vapor pressure and emission
rate volume (cmVsec) were all adjusted for temperature.
                                                                                      Table 4
                                                                     Calculated Wind Speed Constant of Equation 6

                                                                           Clay
                                                          Chemical
                                                                     Wind  Speed mph
                                                                     0.5    2.5    5.0
                                                          Benzene     0.71   0.66  0.71  0.86   0.73  0.74  0.50   0.60    0.65
                                                          CCI<        0.63   0.63  0.71  0.79   0.71  0.72  0.49   0.59    0.67
                                                          Trichlr      0.65   0.65  0.72  0.73   0.73  0.77  0.48   0.61    0.71


                                                            The overall mean for all combinations of soil, chemical and wind
                                                          speed was  0.67.  No  significant difference  in this constant  was
                                                          observed between the chemicals tested. However, a significant  dif-
                                                          ference (p <0.002) was observed between soil types. The difference
                                                          is especially pronounced in the topsoil sample. This implies that in
                                                          using equation 3, the appropriate  wind speed constant should be
                                                          selected depending on the soil t>pe (Table 5).

                                                                                      Tabk5
                                                                               Wind Speed Constants
                                                                                                 Wind Speed Constant
                                                                                                        0.67
                                                                                                        0.75
                                                                                                        0.60
                                                            As expected, the actual wind speed did not correlate well with the
                                                          wind speed constant for the clay and sandy soils. However, topsoil
                                                          did show a strong correlation (R  = 0.97). The reason for this is
                                                          probably due to the organic content of the topsoil sample. The soil
                                                          tested was significantly more organic than either the clay or sand.
                                                          This organic  fraction could have pronounced effects on chemical
                                                          retention through either adsorption or increased porosity.
                                                            Even though Figure 2  shows that the clay soil contained the
                                                          greatest number of small particles,  topsoil was found to have a
                                                          porosity of 51<%,  the highest of all samples. The reason for this is
                                                          the difference between open and trapped pores found in soils. The
                                                          topsoil contained significant amounts of organic detritus, therefore
70
AIR MONITORING

-------
  uch space  existed within particles (i.e.,  porous  wood chips).
   ese trapped air spaces would result in less chemical evaporation
 n. vaPor pore diffusion and would, therefore, explain the reduced
emission rates observed in all topsoil experiments.
  unce the chemical begins to evaporate from the pores of the soil,
 ne actual surface area exposed to the travelling air  decreases
 gnilicantly. Wind speed is not as critical in determining the emis-
fwf wteS ^f below surface contamination.  Therefore, Equation 3
snouid not be used after the material has visibly evaporated from the
son surface. The time for this step to occur is strongly dependent on
wind speed, but for the chemicals and soils tested here, Equation 3
was reasonably accurate for up to 30 min  after a spill.  Since the
wind  speed  constant  determined  in this  experiment  linearly
decreases with  time, Equation 3 could be modified to account for
mis variation. Other equations are available for predicting subsur-
face emission rates.4.' These are also based on Pick's Law and con-
sist of all the variables of Equation 3 as well as soil porosity, tor-
tuosity and the depth of contamination.
  If the equations are to be used for determining the organic vapor
emission rate from open hazardous waste landfills, some modifica-
tions are necessary. First, the experiments presented here are using
moisture-free soil. In reality, all soils contain some moisture which
may inhibit or accelerate evaporation and pore diffusion.  The ef-
fect of moisture on the emission rates can be quite varied depending
on various oil and chemical characteristics.
  Second, the  soil column tested was extremely homogeneous. In
normal situations, the quality of the soil will vary substantially with
depth.
  Third, wind speed is  rarefy consistent, therefore average wind
speeds must be used as well as wind  direction. Finally,  surface
temperature may significantly vary from air temperature at times of
high solar radiation. Because vapor pressure and diffusivity are
both  strongly  affected by temperature,  it is important  to take
temperature  readings at  the air/soil interface  and not rely on am-
bient temperature  readings.
  In  representing  any  environmental  phenomena  using  a
mathematical model, there is a significant chance that the model
either overestimates or underestimates the true  situation.  By in-
troducing new variables in the mathematical relationship and ad-
justing others, reasonable estimates can be made. Previous research
has shown Equation 3 to be within 50% to 150% of the actual emis-
sion rates for all soil and chemical combinations.' By correcting the
relationship of wind speed using the data presented here (Table 5),
the estimates are much closer to the true emission rates  observed.

REFERENCES

1. Cupitt, L.T., Fate of Toxic and Hazardous Materials in the Air Envir-
   onment, USEPA Publication No. 600/S3-80-084, Dec. 1980.
2. Shen, T.T. and Sewell, G.H., "Air Pollution Problems of Uncontrolled
   Hazardous Waste Sites," Proc. National Conf. on Management of Un-
   controlled Hazardous Waste  Sites, Washington, DC, Nov. 1982.
3. Shen, T.T., "Emission Estimation of Hazardous Organic Compounds
   From Waste Disposal Sites,"  73rd Annual Meeting of the Air Pollution
   Control Association, 1980.
4. Thibodeaux,  L.J.,  Chemodynamics:  Environmental Movement  of
   Chemicals in Air, Water and Soil,  John Wiley and Sons Publisher, New
   York, NY, 1979.
5. Treybal, R.E., Mass Transfer Operations,  2nd edition, McGraw-Hill
   Book Company,  New York, NY,  1968.
6. Perry,  R.H. et al.,  Chapter 3: Diffusion Coefficients, Chemical Engi-
   neering Handbook, 5th Edition, McGraw-Hill, New York, NY, 1973.
7. Ziegler, R.C.,  Personal Communication, Calspan  Corporation,  Buf-
   falo, NY, 1979.
8. Arnold, J.H., "Unsteady-State Vaporization and Absorption," Trans-
   action  of American Institute of Chemical Engineers, 40; 1944, 361-379.
9. Caravanos, J., Validation  of Mathematical Models  Predicting the
   Emission Rates of Selected  Organic Solvents from Saturated Soils,
   Doctoral Dissertation, Columbia  University School of Public Health,
   New York, NY, May 1984.
                                                                                                     AIR MONITORING
                                                             71

-------
    AIR MONITORING AT  A  MAJOR  HAZARDOUS WASTE
       CLEANUP  SITE:  OBJECTIVES/STRATEGY/RESULTS
                                          JOHN M. BRUCK
                                      EUGENE W. KOESTERS
                                       WILLIAM R. PARKER
                                     PEDCo Environmental, Inc.
                                           Cincinnati, Ohio
INTRODUCTION
  On December 3, 1982, the U.S. Army Corps of Engineers issued
an invitation for bids for waste removal at the Chem-Dyne haz-
ardous waste site in Hamilton, Ohio. The principal components of
the project included:
•Construct decontamination and drum staging areas
•Confirm or test contents of 30 tanks and 8600 drums
•Dispose of all waste in approved sites by use  of appropriate
 haulers
  The Chem-Dyne site covered approximately 10 acres, and most
of it was enclosed by a fence installed along the perimeter. The site
is bounded on the north by a canal, on the west by a factory build-
ing, on the east by an open field and on the south by a residential
area (Figs. 1 and 2).
  While in operation, the Chem-Dyne facility was used  for the
storage and reprocessing of wastes. At the state of the cleanup ac-
                                              tivity, the site contained 31 above-ground tanks, two below-grade
                                              open-top tanks, approximately 8,600 drums, two tanker-trailers,
                                              eight semitrailers, one flatbed truck and one flatbed railroad car.

                                              AIR MONITORING OBJECTIVES
                                                Among the materials known to be present on the site were large
                                              quantities of organic solvent mixtures. These volatile organic com-
                                              pounds had the potential for producing an  adverse impact on
                                              public health. The  prime cleanup contractor, O.H. Materials,
                                              awarded PEI a contract to conduct perimeter air monitoring at the
                                              site and to perform subsequent (within 48 hr) gas chromatography/
                                              mass spectrometry (GC/MS) analysis of the samples collected. The
                                              objectives of the overall project were  to minimize the release of
                                              any organic vapors and to perform off-site monitoring to measure
                                              concentrations of the various  types of  vapors that may have been
                                              released from the site as a result of the cleanup work and the gen-
                                              eral contaminated nature of the site.
       T
                     OHIO
                            A
                            /u
           ^Hamilton
                   MOTH


     nil inc'Tii* PI«H, IKICI nun t IHOT>
    won >«r> P
    !l!f WT«H!-I
    MT[ DTTAIIS-2
    tit wnmnmie ioc>iim nui
                        cr«i «n>
                        C(I MOTH
                        cm «tn
                                                                                  •Ofl tlH ti
                                                 Figure I
                                          Chem-Dyne Site Location Map
72
AIR MONITORING

-------
                                                                                                              D
>
ye


I
2;
HH
H
o
&
h—i
z
o
                     LEGEND
                •»•* tire •OUMDANV


                —— WORK AHEA


      x         _._ NCW FKMCI



                ^-f--^ CONTAINMENT BCflM



          ^ '      0  FIILD OFFICI• t PAHKIHO


                  ©  DICONTAUINATION AREA



             f     ©  ACCItl KOADWAYI



           *  ©GX5)  TANHAOC APIIAt


      '    "'       0  OCNCHAL WOHKINO AKEA



      i     ,-       ©  »TAa.NO I .TOHAOE A«A



           •
                                                                   J
     :;:;: HEOUIHED LAME WIDTH FOB




'',     (•)  CHEU-DYMC iUlLDlNO


      (d  OflUU CMUlHINO AREA
L _i







FtNCC
•W95
. 5893
.« j
iM9J [~~1' 1
, /
.-,89 < /(
TANKS
1 |


IZH
   ^
                                                                                                       Figure 2

                                                                                             Chem-Dyne Work Area Plan

-------
               Figure 3
Chem-Dyne Ambient Air Sampler Locations

-------
MONITORING

  The ambient air monitoring network for  the  cleanup  opera-
tion at Chem-Dyne consisted  of  one meteorological station and
nine fixed Tenax  ® samplers placed around the perimeter of the
s«e (Hg. 3). The  Tenax®  samplers were pole-mounted, with the
sampler inlets at  an elevation of 10 ft. The meteorological sys-
tem, which measured windspeed, wind direction and temperature,
was installed at an elevation of 33  ft near the observation platform
(outside the perimeter fence). The Tenax® samplers and cartridges
were protected from direct sunlight and precipitation by a highly
polished metal shield. The shield was open on both ends and the
bottom to ensure  unrestricted air flow. PEI had previously devel-
oped this type of sampler for portions of the Love Canal study and
cleanup operations (Fig. 4).
                           Table 1
           Example Meteorological Data and Wind Rose
       TELEDYNE  -  HASTINGS
  MASS FLOW METER 0-100 CCM
  TENAX TUBES
                                       MICRO
                                      NEEDLE
                                       VALVE
                              RUNNING TIME METER

                          Figure 4
             Tenax®  Air Sampling System Diagram

  The flow through each Tenax®  cartridge was controlled by a
micrometer needle valve and was measured before and after each
sampling period with a Hastings mass flowmeter. The average of
the two flow readings was used to determine the air flow rate. The
volume of air sampled was determined by multiplying the flow rate
by time.  Flow was maintained with  a vacuum  pump capable of
pulling a vacuum  in excess of 20 in. of mercury. The performance
of the flowmeter was verified weekly in the PEI Audit System Ver-
ification  Laboratory.  The Tenax®  cartridge was the first com-
ponent in line on  the sampler so the sample would not be in con-
tact with any sample lines or other possible interferences.

Quality Control
  To ensure that the data obtained were representative of the emis-
sions during the cleanup activity, PEI operated all nine samplers
continuously  while cleanup operations were in progress. Opera-
tion  of all nine sites was necessary to allow for changes in weather
conditions. Five samples, one  upwind and four downwind, were
selected for analysis each day after the meteorological data had
been analyzed. Two of the monitoring stations collected two sam-
ples  each to  provide  duplicate samples.  In addition to the five
samples selected for analysis, one of the two sets of collected sam-
ples was analyzed  daily for quality control purposes.
   STATION: CHEH-DYNE
   FILE START DATE: O7/16/B:.
   FILE END DATE: O7/I7/6'
   MUMPER OF DAVS: 2
   DAVS COVERED IN PRINTOUT: 1-2
   WIND   2-7 nF'H
   DIR   HRS    V.
8-11 MPH
HRS    7.
12-25 rlFH
HRS    7.
 >25 flPH
HRS    7.
  TOTAL
HRS   •/.
N
NNE
NE
ENE
E
ESE
SE
SSE
S
SSW
su
usu
w
UNM
NM
NNW

CALM
TOTAL
0
0
0
0
O
0
0
0
0
3 25.
1 6
0
i a.
1 6.
0
0
6 SO.
HOURS
HOURS
.O
. 0
. 0
. 0
. O
. 0
. 0
. 0
. 0
.0
. 3
. 0
.3
.3
. 0
. O
, O


O
0
0
o
0
0
o
o
0
0
0
0
0
o
0
0
o


.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0


0
o
o
0
o
0
0
o
0
o
0
0
0
o
0
0
o


.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0


0
o
o
o
0
o
o
0
0
0
0
o
o
0
0
o
0


.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
-O
.0


o
0
0
0
o
0
0
0
o
3
1
o
1
1
0
0
6
6
12
.0
.0
.0
.0
.0
.0
.0
.0
.0
25.0
B.:
.0
a. j
8.3
.0
.0
50.0
50.0
1OO.O
  The meteorological data (an example of which is shown in Table
1) were delivered  to   PEI  each day  along with the Tenax®
samples. These data were reduced to hourly averages and presented
as a wind rose (example shown in Table 1). The wind rose was then
superimposed on the site map to determine which samples were to
be analyzed.
Analysis

  The field samples were placed in a cooler and delivered to the
PEI laboratory. Upon arrival at the laboratory, they were placed in
a refrigerator, where  they remained until their analysis (within
48 hr of sampling). All Tenax®  tubes were sealed in glass culture
tubes with Teflon® cap liners.  The culture tubes  were placed in
metal cans and stored in a refrigerator until needed. Strict USEPA-
approved chain-of-custody procedures were observed at all times.
  Volatile organics were recovered from the Tenax®  by thermal
desorption and purging with helium into a liquid-nitrogen-cooled
nickel capillary trap. The vapors were then introduced into a high-
resolution glass GC column where the constituents were separated
from each other. Characterization and quantification of the con-
stituents in the sample were accomplished by MS, either by meas-
uring the  intensity of the total ion current signal or by extracted
ion current profile.
  The target compounds specified by the IFB included:
1,2-Dichloroethane
Chloroform
1,1,1-Trichloroethane
Benzene
Carbon tetrachloride
Trichloroethylene
1,1,2-Trichloroethane
Toluene
Perchloroethylene
Chlorolbenzene
Meta-dichlorobenzene
Ethyl benzene
P-xylene
                                                                                                 AIR MONITORING
                                                                                                                           75

-------
                          Table 2
                   Example Analytical Data
STATION:  CHEM-DVNE
FILE START  DATE: O7/I6/BJ
FILE END DATE:  O7/I7/BJ
NUt1E23
MFCS X MRS
0 .0
o .0
I' . o
0 .0
O .0
o .0
o .0
».
(.
c
<.
0
^
C
.0
. 0
.0
.0
.0
.0
.0
o
0
o
o
o
o
o
o
o
o
o
0
0
0
o
1FH
X
.0
.0
.0
.(I
.0
.0
.0
.0
.0
.«'
.0
.0
.0
.0
TOTAL
MRS 7.
O
O
u
O
rt
(>
•''
c.
(I
r.
i
o
i
i
o
0
.(I
.0
.0
.0
Ł,
.0
. o
.0
_ {1
rs.o
B. :
.0
a. :
B. J
.0
.0
 CALrt HOURS


 TOTAL HOURS
                                             6  3O.O

                                             o  10.0

                                            I?  IOO.O
  Analytical quality control procedures were followed at all times.
These included spiked tubes  and blanks and  instrument tuning
according to 1FB specifications and the procedures outlined in the
PE1 Laboratory Quality Assurance Plan.
  After completing the analyses, PEI delivered written reports to
the prime contractor's Quality  Assurance Supervisor during the
next visit to the site. Results were reported in mg/m1 (ng/ml) for
easy comparison with published TLV tables. If a value greater than
10% of the TLV  was observed, the prime contractor's Quality
Assurance Supervisor was notified by telephone.
  Except  for occasional  electrical current interruption  or pump
malfunction (quickly remedied with spare parts by PEI's on-site
technician),  the project  proceeded smoothly for more  than 120
work days (approximately 180 calendar days). Some concern was
expressed over "channeling" of air currents due to the proximity of
two samplers to long,  multistory buildings. Some alternative loca-
tions were proposed.  Since none offered better conditions, the
samplers remained where originally placed.
   When two security guards became ill one night while the site was
inactive, round-the-clock sampling was proposed. It was deter-
mined, however, that  the episode was unrelated to emissions from
the site and sampling continued to be conducted during working
hours only.
                                           U i r. d  R o s
                              J
                                                         CONCLUSIONS
                                                           An example of the analytical results is shown in Table 2. Not
                                                         once during the  entire project did analyses indicate that air levels
                                                         of any of the target compounds exceeded 10% of the applicable
                                                         TLV. Even though levels of target and other compounds remained
                                                         low relative to their respective TLVs,  this  monitoring  technique
                                                         does not replace direct measurements by OVA or  photometric
                                                         methods which are critical to on-site personnel health and safety
                                                         management.
76
AIR MONITORING

-------
       MEASUREMENT  OF VOLATILE ORGANIC EMISSIONS
                      FROM  SUBSURFACE CONTAMINANTS

                                             W. DAVID BALFOUR
                                              BART M. EKLUND
                                               Radian Corporation
                                                   Austin, Texas
                                          SHELLY J. WILLIAMSON
                                    U.S. Environmental Protection Agency
                                                Las  Vegas, Nevada
INTRODUCTION

  Organic subsurface contaminants present a potential for human
exposure via the pathways of both air and water. The migration of
volatile  organic compounds (VOCs)  from  contaminated soils or
groundwater into the air represents a potentially major source of
exposure which has not yet been adequately assessed.
  To assess the exposure potential of this  pathway, a method is
needed  to directly  measure gas emission rates from  the con-
taminated  material or overlying soil. It should be emphasized that
the need is for a measure of the gas emission rate, not just a gas
concentration.
  The emission rate data would permit an exposure assessment
through the use of existing models and define the need to control
gas emissions from subsurface contamination sites or at hazardous
waste facilities. Application of this measurement method may in-
clude calculation of losses of volatile organic compounds from
storage  tanks,  pipelines, surface spills and/or ponds as well as
direct emissions from surface impoundments, land treatment sites
and landfills.
  The USEPA Environmental Monitoring Systems Laboratory has
been sponsoring research to select, document and demonstrate an
appropriate method for directly measuring gas emission rates from
contaminated soils and/or groundwater for exposure assessment
purposes.  Based upon a review of  the relevant  literature, the
enclosure emission  measurement technique was selected as the
method  of choice. A draft  protocol  for the technique  has been
completed. It presents the principle of the method, descriptions of
the equipment, calibration procedures, quality assurance and quali-
ty control procedures and an operational protocol.
  To date,  the method has been demonstrated at several field sites.
Parametric tests were performed to evaluate the effect of specific
variables on the measured emission rates. In the second phase of
the study,  laboratory studies have investigated the effect of en-
vironmental and operating variables on the volatilization and emis-
sion rate of organic compounds in order to verify the validity of the
method. The end product of this research will be a protocol for use
in measuring volatile organic compound emission rates from con-
taminated soils  and/or groundwater.
  This paper presents the results of field measurements  at a spill
site. Included are an assessment of emissions  from the site,  an
evaluation  of the effect of specific  operating variables on the
measured emission rates, an evaluation of the appropriateness of
the statistical sampling procedure for area sources and an analysis
of the sources of variability in the method.
DESCRIPTION OF METHOD

  The emission isolation flux chamber is a device used to make a
direct emission measurement. The enclosure approach has been
used by researchers to measure emission fluxes of sulfur and
volatile organic species.l'2>3 The approach uses an enclosure device
(flux chamber) to sample gaseous emissions from a defined surface
area. Clean dry sweep air is added to the chamber at a fixed con-
trolled rate. The volumetric flow rate of sweep air through the
chamber is recorded, and the concentration of the species of in-
terest is measured at the exit of the chamber. The emission rate is
expressed as:
      E; = CjR/A
                                           (1)
where
  Q

  R
  A
emission rate of component i, /ig/m2-sec
concentration of component i in the air flowing from
the chamber, /ig/m3
flow rate of air through the chamber, mVsec
surface area enclosed by the chamber, m2
All parameters in Equation 1 are measured directly.
  A diagram of the flux chamber apparatus used for measuring
emission rates is shown in Figure 1. The sampling equipment con-
sists of a stainless steel/acrylic chamber with impeller, ultra high
purity sweep air  and rotameter  for measuring  flow into the
chamber and a sampling manifold for monitoring and/or collection
of the specie(s) of interest. Concentrations of total hydrocarbons
are monitored continuously in the chamber outlet gas stream using
a portable flame ionization detector (FID) and/or photoionization
                                       SAMPLE COLLECTION
                                        ANDIOH ANALYSIS
                                                ON/OFF FLOW
                                                  CONTROL
                                                GRAB SAMPLE
                                                  PORT
                                          STAINLESS
                                          STEEL COLLAR
                       Figure 1
     Cutaway Side View of Emission Isolation Flux Chamber
                 and Sampling Apparatus
                                                                                         AIR MONITORING
                                                                                                                 77

-------
detector (PID)-based analyzers. Samples are collected for subse-
quent gas chromatographic (GC) analysis once a steady-state emis-
sion rate is obtained. Air and soil/liquid temperatures are measured
using a thermocouple.
  To determine the emission rate for a source of much greater area
than  that isolated, by the  flux chamber, a sufficient  number  of
measurements  must  be  taken at  different locations  to  provide
statistical confidence limits for the mean emission rate. The area
sources measured were gridded (30 ft x  75 ft) and a minimum of
six measurements (when possible) to account for spatial variability.
Additionally, a single point was selected as a control point to define
temporal variability. On-site OC analyses  were performed for  all
flux chamber measurements, and several canister samples were col-
lected for each area to allow off-site detailed GC analysis.  Prior to
using the chamber, blank and species recovery data were obtained.

FIELD TEST

   The field test took place at  a Marine Corps helicopter facility in
Tustin, California. The spill site was an abandoned practice fire-
fighting area where JP-4 aviation fuel had permeated the local soil
resulting in a hydrocarbon lens at the surface of the saturated zone.
Much of the contaminated surface soil has  been excavated and
backfilled  with sand. The estimated area of the  contaminated
groundwater plume is 45,000  ft*. The overlying soil in this area is
primarily sandy silt with a large clay fraction. Depth  to  ground-
water is nominally  16 ft below land surface. Product was being
recovered  from  the groundwater in  the area at  a  rate of  50
gal/month.
   Field tests involved gridding the area  to be measured,  dividing
the gridded area into two distinct zones (upper and lower) and con-
ducting the flux chamber measurements at randomly selected grid-
points. Following the first set  of flux chamber measurements, a se-
cond  set  of measurements was made to evaluate  six variables
associated  with the flux chamber design and/or operation. The
variables assessed during this parametric study are listed in Table 1.
The approach used to conduct the parametric  study was to make a
measurement at the  baseline  conditions, followed by  a measure-
ment at the modified condition. A set of six paired  measurements
was typically made at a number of different gridpoints.

                             Table 1
            Parametric Test Studies of the Flux Chamber
                        Equipment or
                    Procedural Kodif LCCCIOD
                                        fteeult
                                     of Modification
Chamber geometry
Chember opacity
          Hie of fUc topped, 23L
          cbember u piece of typi-
          cal domed, 31L dumber

          Chember covered vich 4—5
          leyer* of cbia. black
          polyethylene «beecui(
                                           Chember geometry verie4
dumber Cempe
                   aeetini coil vicb lemper-
                   ecure controller pieced
                   iaaide cbember
                                           Chemb
                                           Cbemb
                                                r blocked from
                                                Ler redietion
                                       r maintained et
                                       bove typical
sULettivtt tumidity
Sweep air flow rate
Sweep «ir pa*a*d through
•a ij.pia|«r of DZ water
Sweep *ir flow r*te de-
tumidity iacreasad
Sweep air flow rate
1/3 of baceliiie rate
(1.1 L/Bifl v*. 3.2
L/.u.)
Impellor rate
                   Chember operated with
                   impellor off
                                  KLciag cherecteria
                                  of chamber veried
   A statistical analysis of the data was performed to investigate the
 effect of the variables (chamber geometry, impeller rate, chamber
 opacity, relative humidity, sweep air flow rate and temperature) on
 the VOC emission rates measured in the field.  Emission rates were
 calculated  from the sampling and analytical data as shown above
 (Eq.  1).
                                                                                      Table 2
                                                              Comparison of Air and Liquid Samples with JP-4 Composition
                                                           PrlMry Component!   Liquid Sample   Beftdtpace Sample   FluK Clumber Semple
                                                              of J?-4*          Well 121.°       Well »24b        Grid Point 21°
o-r.otan.
2/3-melhylpentaa<
n-bexane
metbylcyclopcntaor
eye lobeiane
2/3-metbylbexaoe
n-hept ant
metby Icyc lobeiaoe
toluene
2/3-methylheptaoe
e-octane
o-iyleo*
o-nonane
1,2.3 tn»etbyl-
beoteoe
o-d,cane



X
I
I
I
X
I
X
I
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
I
X
X
X
X

X
X
I
I
I
X
I
X
X
X
X
X
                                                           n-d odcce.ae

                                                           a-tridccaoc
                                                          'Compound! preient at greater than 1  percent by wight  La J?-4
                                                           fuel
                                                          °Anal»n.  by capillary CC-FID/PID
  A paired t-test  was used to test  the  hypothesis that a given
variable  influenced  the measured  emission  rate.  Analysis of
covariance was used to account for the fact that the temperature in
the chamber varied from one measurement to the next. This treat-
ment made it possible to differentiate between the random scatter
and temperature trends in  the paired measurement data. In this
manner,  the error variance in the random scatter was minimized
making it easier to identify the effects of the variables of interest.

DISCUSSION OF RESULTS

  The  compounds identified  during the emission measurements
were compared against the major components of the JP-4 aviation
fuel, the contaminant layer and the vapor space in one of the ex-
ploratory wells. As shown  in  Table  2, the major components of
JP-4 aviation  fuel were identified in the contaminant  layer,  well
vapor space and emissions  measured with the flux chamber. This
fingerprint indicates that the emissions measured were the result of
the aviation fuel.
  Air temperatures in the flux chamber varied over the course of a
day, with changes in the ambient temperature  (90°-115°F).  The
emission rates measured at a single location were correlated with
the air  temperature within the flux chamber over the course of day.
The observed trends in the data are shown in Figure 2. This correla-
tion was  used to  correct the  measured  emission rates to values
which would be expected if all measurements had been made at an
average air temperature within the chamber.
  There  was  typically   a  difference  between  the  ambient  air
temperature  and  the  air  temperatures measured  within  the
chamber. This difference typically ranged from 2° to 18TF. Soil
temperatures were difficult to measure accurately; however,  the
values measured indicated minimal differences (C4°F) between the
inside and outside of the chamber. Soil temperatures did vary over
the course of the day but to a lesser extent than the air temperature.
The variation in soil temperature was typically less than 9°F.
78
AIR MONITORING

-------
50
40

O
20
                                                           400
300
                120     180      240     300     360     420
                       Sampling Time (min.)                  3 p.m.

                           Figure 2
     Variations in Temperature and Total Hydrocarbon vs. Time
                  at a Single Sampling Location


line lop
Cl.—Wr


lln


t la
Of


r

'l
line Clu-bt
r (
{ I

r lln* Sweep
Air
j
' r
1
1
""* ;~~
,,

line CI>»1>«
                           Figure 3
             Mean Emission Rates per Variable Tested
   The mean corrected emission rates for each parameter tested are
 shown in Figure 3 with the corresponding baseline data. The error
 bars represent the 95% confidence interval. The large error bars
 present for the impeller rate tests are due to only two tests being
 performed for this variable versus six tests for each of the other
 variables. The data shown in Figure 3 are presented in Table 3.
   The statistical significance of each parameter  is presented in
 Table 4. At the 95% probability level, only opacity and sweep air
 flow rate were significant. At the 60%  probability level, all
 variables except chamber geometry were significant. No statistical
 significance means that there was not enough evidence in the data
 to conclude a statistical significance.
   Two different flux chamber geometries were tested, one having a
 flat acrylic top and the other (baseline condition) a domed acrylic
 top. Either of these designs is expected to be suitable, with the use
 of an impeller for mixing. Tests run without the impeller showed a
 decrease  in the measured emission rate. Additionally, a greater
 amount of variability was observed in the concentration levels in
 real-time. For these reasons, it is suggested that an impeller be used
 to assure proper mixing for any given chamber geometry.
   Significant differences in the measured emission rate were noted
 when incoming sunlight was shielded from the chamber. Internal
 air  temperatures were also lowered; however, the change in air
 temperature does not  account  for the total effect observed. As
previously stated, the soil temperatures were difficult to measure.
The shielded sunlight may cause a change in soil temperature at the
surface  which could not  be measured.  This effect will  be in-
vestigated in future studies.
  The relative humidity of the air inside the flux chamber will vary
depending upon the soil conditions. Baseline operation of the flux
chamber is with dry sweep air. Relative humidities from  10 to 95%
have been observed.  Condensation can form in the chamber  at
relative humidities greater than 70%.  During the tests, an increase
in relative humidity to 60% was obtained  by passing the sweep air
through a bubbler. No condensation was observed in the chamber
at these relative humidities, and no effect on the measured emission
rate was noted. Dry  sweep air whould be used for flux chamber
operation.
  The flux chamber is routinely operated at sweep air flow rates of
3-5 1/min. The flow rate can be lowered  when emission rates are
low in an attempt to reduce the dilution and allow analytical detec-
tion. For these tests, the flow  rate was reduced to 1 1/min. This
change significantly reduced the emission rate measured. At this
time,  it is not known if this effect is the result of changing the air
phase concentration or the mixing at the air-soil interface. This ef-
fect   does  have  important  implications  for  the  sampling
methodology and is being investigated in  Phases II and III of the
project.
  The chamber  air  temperature showed significant trends  with
emission rate for the eight-hour test (as  previously shown), was
significant at the 60% probability level for the parametric test, but
was not significant at the 95% probability level. It is expected that
temperature can  have an  effect on  the measured emission rate.
However, the air temperature will be a variable which  cannot be
controlled. For this  reason, it  is important that  temperature be
monitored during measurements. Additional studies will investigate
the dependence of the emission process on temperture  so that an
appropriate comparison can be made between measurements at dif-
ferent temperatures.
  The variability in the measured emission rates was due to a varie-
ty of factors, as shown in Table 5. Only one-third of the variability
was due to sampling and analytical considerations, i.e., under the
control  of the experimenter. As shown, 17% of the total sampling
variability was attributed to spatial considerations while 24% of the
total  variability  was attributed to temporal considerations. As
shown,  25% of the total variance was eliminated by correcting for
temperature. This supports the fact that temperature is a  significant
variable.
  One of the objectives of this field  demonstration  was to deter-
mine  if the average of six individual  flux chamber measurements
adequately represents the emission rate from an entire area source.
As a test of this hypothesis, the average emission rate from various
subsets of six points was compared to the  overall average emission
rate. Emission data from a single day were randomly selected for
each of the 12 gridpoints measured. The total variability  of the ran-
domly chosen data includes spatial,  temporal and  sampling/an-
alytical variability (data were corrected to an average temperature).
The summary statistics for the 12 values chosen were as follows:

    overall mean Oig-C/m2-sec)	44.6
    standard deviation (/tg-C/m2-sec)	15.9
    coefficient of variation (%)	35.5

All possible subsets of size six were chosen from  the 12 adjusted
emission rate values; means, standard deviations and coefficients
of variations were computed for each. Of the 924 subset means,
99.8% were within 25% of the overall mean and 66% were within
10% of the overall mean. The  overall mean was  within the  95%
confidence intervals of 99.8% of the subset means. Coefficients of
variations ranged from 13% to 51% for the subsets.  Thus, the
authors believe that the procedure was adequate in calculating an
average emission rate representative of the entire area.
  The overall emission rate estimate must  not be confused with the
population emission rate.  A  population emission  rate  average
                                                                                                   AIR MONITORING
                                                                                                                             79

-------
                                                                   Table 3
                                      Analysis of Covariance Results for Examining the Effects of Varioiu
                                                      Factors on Measured Emission Rate



tape 11 or Kate

Opacity

ke la live
Humidity
Sweep Air
Klow Kate
Temperature

Teat
Condition

Flattop Chamber
Baaellne
Upellor Off
Baaellne
Opaque Chamber
Baaellne
Sat'd Sweap Air
Baeellna
Low Flow Bate
Baaellne
lleatvd Chamber
No. of
Data
Polnta
a
6
2
2
6
t
6
5
7

6
t
Ealaalon Rate
Adjuated
He an*
J6.3
32.4
54. •
**••
31.1
21.2
60.4
6S.I
49.4

1!.9
64. t
Std. Statistical
Error Sl»nlf Ic«nc«b
3 	
.m Uo
1.01
4
4
2
2
2
3
4

4
4
.59 No
.39
.19 Yea
.19
.74 No
.10
.49 Yea

.70 No
.70
laaldual
Std. Day.
(ua.-C/ar'-Mc)' «*(!)"
7.17 12.7

t.iO 96.0

5.65 91.1

6.62 95.9

11.71 12.4

11.5 91.1

 ». With the exception of the temperature test series d«la, mean emission rates were corrected for
   the concomitant variable (covariate) of average air temperature in the chamber.
 b. Based on the analysis of covariance results, does the lest condition have a significant effect on
   measured emission rate at the 93** probability level a - 0.05)7
 c. The residual standard deviation provides an estimate of emission rate measurement repeal-
   ability (a measure of differences that are likely to occur between repeal measurements of the
                                                                 same (rid point under the same conditoru) This U an indirect estimate of repeaiablity thai
                                                                 depends  on the accuracy of the model (error* in Uie model tend to  increase the residual
                                                                 standard deviation). A better way to estimate repeaiabtity. therefore, is to do repeat tests.
                                                               d The R1 statistic measure) the amount of variation in emission rate which has been explained by
                                                                 the terms in the model (jrid point, lest condition, and average temperature in the chamber).
                              Table 4
       Significance Levels of Tustin Lower Level Sampling Data
                                                                                              Table 5
                                                                     Components of Variability for Fin Chamber Measurements
Test Series
Chamber Geometry
Impeller Rate
Opacity
Relative Humidity
Sweep Air Flow
Rate
Temperature
T-Valne*
0.907
1.542
3.401
-1.24
3.797
1.000
Probability
of Greater T|
0.416
0.366
0.027
0.303
0.013
0.364
Statistical
Significance}
No
No
Yes
No
Yes
No
* The student's T value for testing the hypothesis thai the test condition (baseline vs. non-baseline)
  does not have a significant effect on emission rate (adjusted for temperature).
t Probability of a greater absolute T value assuming adjusted emission rate differences follow a
  normal distribution.
t Based on Ihe analysis of covanance results, does the test condition have a significant effect,on
  measured emission rale at the 95^ probability level (^= 0.05)7 Given a conclusion of a statistical
  difference, there is a 5% probability that we are wrong.

would  require  complete  continuous measurements with time over
the entire area. This, of course, is not  possible. For this reason,
probability  (random) sampling in  space and time  are employed
when determining the average emission rate  values.
  The average  emission rate measured at the spill site was 0.24 jtg-
C/m2-sec  for the  upper  level and  41.6 /ig-C/m^-sec  for the lower
level. These emission rates correspond to a daily emission rate from
the areas of 0.18 kg-C/day. Extrapolated  to a yearly value, this is
an emission rate of 65  kg-C/year of total hydrocarbons.

CONCLUSIONS

  The  flux  chamber  field  tests have  provided  insights  to the
variables which affect the  measurement  method, the variability
associated with the sampling procedure and the appropriateness  of
the statistical sampling procedure for area spills.  Specifically:
•Measured emission rates do vary with  air  temperature, and it is
  possible to adjust the values to an average temperature, thereby
  lowering the variability associated with the measurements
•Measured emission rates decreased when sweep air  flow rate was
  decreased
•Measured  emission rates decreased when  sunlight  was shielded
  from the chamber
•Measured emission  rates were  not affected by changes in chamber
  geometry and relative humidity
                                                                          Source of
                                                                          Variability
                                                                                         V aria net
                                                                                       Conpoaral*
«i of Total
 Vartucc
•* of Total
MiDuaTem-
  perMnre
  Virtue*
                                                                Temperature in the
                                                                  Chamber                 132.5
                                                                Grid (Sampling
                                                                  Location)                 86.7
                                                                Temporal (Day-to-Day)       121.0
                                                                Sampling/Analytical          163.8
                                                                       TOTAL            504.0
   26.3

   17.2
   24.0
   32.5

   100.0
   23.3
   32.6
   44.1
   100.0
                                                                • Variance components are equal to standard deviations squared and thus have units which are Ihe
                                                                 squares of that used for the measured emissKm rates (Mf-Om^-sec).1
                                                                •The sampling and analytical variability associated with the  flux
                                                                 chamber measurements are estimated at 33%
                                                                •The statistical  sampling approach does appear to result in an
                                                                 average emission rate representative of the entire area
                                                                  These findings are not considered to be definitive but will  pro-
                                                                vide the direction for  additional study.

                                                                DISCLAIMER

                                                                  Although the research described in this paper has been funded in
                                                                part by  the USEPA through Contract Number 68-02-3513 to Ra-
                                                                dian Corporation, it has not been subjected to Agency review and
                                                                therefore does not  necessarily reflect the views of the Agency and
                                                                no official endorsement should be inferred.

                                                                REFERENCES

                                                                1.  Cox, R.D., Baughman, K.J. and Earp, R.F., "A Generalized Screen-
                                                                   ing and Analysis  Procedure for Organic  Emissions from Hazardous
                                                                   Waste Disposal Sites." Proc. of the National Conference and Exhibi-
                                                                   tion on Management of Uncontrolled Waste Sites, Washington, DC,
                                                                   1982,  58-62.
                                                                2.  Hill,  F.B., Aneja, V.P. and Felder, R.M., "A Technique for Measure-
                                                                   ments of Biogenic Sulfur Emission Fluxes. J. Environ.  Sci.  Health
                                                                   (AIB(3),  1978, 199-225.
                                                                3.  Adams, D.F.,  Pack, M.R., Bamesberger, W.L. and Sherrard,  A.E.
                                                                   "Measurement  of Biogenic Sulfur-Containing Gas  Emissions  from
                                                                   Soils and Vegetation." Proc. of 71st Annual APCA Meeting, Houston,
                                                                   TX, 1978, 76-78.
80
AIR MONITORING

-------
  ON-SITE AIR MONITORING CLASSIFICATION  BY THE USE
                    OF A TWO-STAGE  COLLECTION TUBE
                                           RODNEY D. TURPIN
                                  U.S. Environmental Protection Agency
                                       Environmental Response Team
                                             Edison, New Jersey
                                              KIRIT H. VORA
                                 Clayton Environmental Consultants,  Inc.
                                             Edison, New Jersey
                                                  J. SINGH
                                               A.W. EISSLER
                                         DARYL STRANDBERGH
                                 Clayton Environmental Consultants,  Inc.
                                            Southfield, Michigan
INTRODUCTION

  The  USEPA's  Environmental Response Team (ERT)  was
established in October, 1978  to provide technical assistance to
Federal On-Scene  Coordinators (OSC), Regional Response Team
(RRT), National Response Team (NRT), USEPA Headquarters/
Regional Offices and other government agencies in the area of en-
vironmental emergency issues  such as chemical spills and uncon-
trolled hazardous waste sites.
  In this paper, the authors describe the two-stage air sampling
tube developed jointly by ERT and Oil and Hazardous Materials
Spills Branch, Edison, New Jersey. The project was based on state-
of-the-art techniques with the objective of developing a convenient
screening media for air samples at sites where unknown  and multi-
ple contaminants may be present. The authors describe the tube
development, sampling rates and method of analysis.
PROJECT SCOPE OF WORK

  Imagine  a 20-acre site anywhere in the country with 5,000 to
10,000 unidentified 55-gal drums and four unlined waste lagoons
containing unidentified liquids  adjacent to a housing development.
You have been asked to conduct an emergency air monitoring pro-
gram. What  collection medium  would you select?  Charcoal?
Tenax-GC? Florisil? Silica gel? What sample flowrate would  you
use (10 cc/min, 50 cc/min, 100 cc/min. 1 1/min, 2 i/min)? What
would be the appropriate sampling volume? What is the minimum
analytical turnaround time? What is the appropriate analytical pro-
tocol? What would be the total number of samples collected per
station? These are just a few of the questions one would ask oneself
if given this assignment. Obviously,  the assignment becomes less
difficult with more information to evaluate. A multistage tube was
developed to provide a quick profile of compounds encountered at
a typical hazardous waste site.
  The fastest  way to get a  rapid  qualitative and quantitative
characterization of an unknown mixture is by gas chromatography/
mass spectrometry (GC/MS).  Solid sorbent  media are  most con-
venient for field work. Thermal desorption with cryogenic trapping
requires minimal sample preparation and permits a very "clean"
sample injection onto  the GC/MS analytical equipment.
  To minimize the cost and development period, the project  was
originally restricted to three candidate tube configurations, each
having three separate stages:
•Candidate Tube  A—To consist  of a polyurethane foam first
 stage, Tenax-GC second stage and an activated carbon third
 stage
•Candidate Tube B—To  consist of a polyurethane foam first
 stage, Chromosorb  102 second stage and an activated carbon
 third stage
•Candidate Tube C— To  consist of a polyurethane foam first
 stage, Porapak S second stage and an activated carbon third
 stage
EVALUATION PROCESS
  The evaluation was made in three phases. First, a preliminary
study  was conducted  to  ascertain the thermal  desorption
characteristics of the candidate sampling media and to optimize the
thermal desorber and GC/MS conditions. The results of  the
preliminary investigation  indicated that  polyurethane  foam,
Porapak S and activated charcoal were not suitable media for ther-
mal desorption. These media were excluded from further study.
The candidate media not rejected in Phase 1 (Chromosorb-102 and
Tenax-GC) were spiked directly with a liquid solution of  the
chemicals shown in Table  1.
                        Table 1
           Liquid Solution Used for Media Evaluation
Material
Isopropyl alcohol
Methyl isobutyl ketone
Ethylbenzene
Aniline
2,4-Dichlorophenol
Naphthalene
Chlordane
Aroclor 1254
n-Nitrosodimethylamine
bis(2-Chloroethyl) ether
Tricresyl phosphate (Technical)
Quantity
(M!)
39
16
17
1.0
0.66
1.0
1.0
1.0
0.067
0.066
0.60
% of TLV for
a 1-1 Air
Sample
4.0
7.8
3.9
10
—
2.0
200
200
—
—
600'
                                                          •Based on TLV for triorthocresyl phosphate (0.1 mg/ra3)


                                                                                       AIR MONITORING      81

-------
The results of the spiking study are summarized in Tables 2-5.
                            Table 2
                    Spiking Study—Tenax-GC
                (% Recovered ft 250°C Desorption)
Spike 1 Spike 1 Spike 3 Mem
Isopropyl alcohol
Methyl isobutyl ketone
Ethylbenzene
Aniline
2,4-Dichlorophenol
Naphthalene
Chlordane
Arochlor 1254

Spiking
55 60 64
79 82 89
95 95 83
68 70 88
100 160 180
51 79 88
62 88 100
87 88 110
Table 3
Study— Chromosorb 102
60
83
91
75
150
73
83
95


(*h Recovered with 150°C Desorption)
Material
Isopropyl alcohol
Methyl isobutyl ketone
Ethylbenzene
Aniline
2,4-Dichlorophenol
Naphthalene
Chlordane
Arochlor 1254
Spike 1 Spike 2 Spike
96 84 88
92 110 95
84 98 76
23 34 46
37 53 27
33 24 7
ND« 1 1 5
ND« 10 5
3 Mean
89
99
86
34
39
21
5
5
-Non-detectable

   A separate spiking study was conducted for n-nitrosodimethyl-
amine and bis (2-chloroethyl) ether. In this study, 100 mg sections
of Tenax-GC and 200 mg sections of Chromosorb 102 were spiked
with  1.0 id  of a  dodecane  solution containing  0.844  mg  of
n-nitrosodimethylamine  and   1.22 mg  of  bis  (2-chloroethyl)
ether/ml of solution. The resultant spike levels were 0.844 and 1.22
fig, respectively. The results of the spiking  stidy  indicate good
recovery from Tenax-GC,  and  somewhat excessive recovery from
Chromosorb  102 (Tables 4 and 5).

                             Table  4
                     Spiking Study—Tenax-GC
                (%  Recovered at 250° C Desorption)
Material

n-Nitrosodimethybunine


bis (2-Chloroethyl) ether
                           Spike 1 Spike 2 Spike 3 Mean**

                            99    94    39     92
                                       (void)
                           120    %    43     110
                                       (void)
••Excluding Spike 3, which was voided due to determinate error (cryogenic trap liquid nitrogen level
was low).

                             Table 5
                  Spiking Study—Chromojorb 102
                (%  Recovered at  150° C Desorption)
Material

n-Nitrosodimethylamine
bis (2-Chloroethyl) ether
                           Spike 1 Spike 2 Spike 3 Mean
                           200
                           240
330
310
340
330
290
290
  The results of the spiking study indicated that the Tenax-GC sec-
tion should precede the Chromosorb 102 section in the final tube
configuration.
  Based on the results from Phase 2, a tube was selected for further
evaluation. This evaluation consisted of generating known concen-
trations of the same materials at different relative humidities (20, 50
                                                            and 90%)  in  a stainless  steel calibration  chamber  at  25°  C.
                                                            Duplicate samples were collected at three flow rates (10, 20 and SO
                                                            cc/min)  and  analyzed  by  thermal   desorption/cryogenic
                                                            trapping/GC-MS to evaluate overall recovery for the sampling and
                                                            analytical system. The following six tables summarize the results of
                                                            the Phase 3 testing:
                                                                                         Table 6
                                                                          Total Recovery at W!t Relative Humidity
                                                                               (ft of Actual Concentration)
Nominal
Flowrale:
Isopropyl alcohol
Methyl Uobulyl ketone
Ethylbenzene
Aniline
2-4-Dlchlorophenol
Naphthalene

Tool

Nominal
Flowrale
Isopropyt alcohol
Melhyl isobutyl ketone
Elhylbenxene
Aniline
2.4-Dichlorophenol
Naphthalene
10 ml/mm 20 ml/mm 50 ml/ nun
Tube 1 Tube 2 Tube 1 Tube 2 Tube 1
170 210 82 73 130
170 190 97 96 180
2W IN 150 170 IK)
160 310 200 210 180
4.9 84 56 87 15
110 130 100 110 130
Table?
Recovery al SO^t Relative Humidity
(ft of Actual Concentration)
10 ml mil 2* ad/all Mai
Tube 1 Tibe 2 Tibc I Take 2 Tike 1
Void* 140 240 170 190
Void 130 110 130 M
Void 1)0 140 140 120
Void 32 21 21 23
Void 100 149 150 63
Void 89 ~9 89 85
Tube 2
200
160
250
110
Oi
120



1 nil
Tikel
190
100
140
6
11
no
                                                            •Sample voided—faulty vaNe on umpling pump prevented drawing air through tube


                                                                                        Tablet
                                                                         Total Recovery at 90^i Relative Humidity
                                                                              (*h of Actual Concentration)
Nominal
Floirratr
Isopropyl alcohol
Methyl isobutyl ketone
Ethylbenzene
Aniline
2.4-Dichlorophenol
Naphthalene
10 ml/mia 20 mL'mim
Tibet
Void'
Void
Void
Void
Void
Void
Tike I
41
120
110
86
150
64
Tikel
44
79
120
15
16
56
Tikel
90
92
130
6.0
7
-3
Mai Bin
Tikel
46
97
110
33
82
60
Tike 2
41
100
100
49
}4
69
                           •Sample voided—faulty valve on sampling pump prevented drawing air through tube.

                              A  separate calibration chamber test was conducted  in which
                           tubes were challenged  at  50%  relative  humidity  only  with
                           n-nitrosodimethylamine  (challenge concentration: 0.844 mg/m3)
                           and  bis  (2-chloroethyl) ether  (challenge  concentration:  1.22
                           mg/m3).

                                                       Table 9
                                 Recovery from Tenax-GC Section at 50% Relative Humidity
                                   (% of Challenge Concentration  Indicated by Sample)
Nominal 10 ml/mln 34 ml/mln $0 ml
Flowrale Tube 1 Tnbe 2 Tibe 1 Tube 1 Tub* 1
n-Nilrosodimethytamine 100 97 83 Void* 34
bis (2-Chloroethyl) ether 93 120 84 Void 100
rain
Tube}
79
91
                                                            •Sample voided—faulty valve on sampling pump presented air being drawn through tube.

                                                              Neither material was detected in any of the Chromosorb  102 sec-
                                                            tions.
                                                              Because Arochlor 1254 and chlordane failed to volatize properly
                                                            in  the  stainless  steel  chamber,   separate  tests  were  run  by
82
AIR MONITORING

-------
evaporating those materials from a glass wool plug into an ambient
humidity airstream. "Challenge" concentrations were measured by
simultaneous sampling of the streams by NIOSH Methods S-244
and S-278:

                          Table 10
          Arochlor 1254 Recovery from Tenax-GC Section
                   at 54% Relative Humidity
Sample
Number
TCA-1

TCA-2

TCA-3*

Sampling
Period
(mln)
100

100

100

Sample
Volume
0)
4.76

4.82

4.51

Challenge

-------
Analysis of Samples
Thermal Desorption of the Tenax-GC Section
  Score the tube in the center with a sharp file and break in half.
Cap the exposed end of the Chromosorb 102 section and return to
storage for later analysis. Scribe the flame-sealed end of the Tenax-
GC section 1 cm from the end of the taper (on the cylindrical por-
tion) and  break  evenly. With the  trap in liquid nitrogen and the
valve in the thermal desorption position, connect one end of the
tube by means of an 0.8 cm Swagelok fitting with Teflon ferrule to
the heated line entering the valve assembly. Again, using an 0.8 cm
Swagelok  fitting with Teflon  ferrule, connect the helium purge line
to the other end of the tube. The helium flow should be 30 ml/min.
Open the  desorber block (preheated  to 250° C), place the tube in
the channel, then close the block. Desorb the sample for 10 min at
250° C before rotating the valve, removing the liquid nitrogen bath
and turning on the trap heater. The carrier gas flowrate should be
in the range of 1 to 2 ml/min depending on the capillary column.
Analysis of Chromosorb 102 Section
   The analysis is identical to that  described for the above, except
the thermal desorber block temperature is 150° C.
Preparation of Standards
   Depending on the findings  of the sample analysis, standards may
be run to confirm  identifications and to quantify  the  materials
found. Due to the small quantities required  (0.1  to 1.0 /i), the
material(s) of  interest must  be handled in solutions. The solvent
                                                        used to prepare the standards must be selected with care to ensure
                                                        that it does not interfere with the GC-MS determination. Prior to
                                                        injecting an aliquot of the standard solution into a glass wool plug,
                                                        the 8 mm O.D., 11 cm long tube holding the glass wool is chilled
                                                        (frosted) over liquid nitrogen to minimize evaporation of volatile
                                                        components while the Swagelok connections are being made.
                                                          As an alternative to the chilling, a septum on the helium inlet fit-
                                                        ting may be constructed  to permit injection of the  standard after
                                                        the Swagelok connections have been made.

                                                        CONCLUSIONS
                                                          If the two-stage tube continues to live up to the evaluation study,
                                                        it appears that the USEPA will be able to collect fewer air samples,
                                                        receive better turnaround time and obtain a sufficient identification
                                                        of many of the low-level organic air contaminants. With this infor-
                                                        mation, the USEPA will be better able to identify  and develop a
                                                        site-specific air monitoring program.

                                                        ACKNOWLEDGEMENT
                                                          The authors express their appreciation to all members of the
                                                        USEPA's Environmental Response Team for their many contribu-
                                                        tions and constant updating, and their indebtedness to all members
                                                        of the USEPA's Oil  and  Hazardous  Material  Spills  Branch,
                                                        Edison, New Jersey, as well as ERT-TAT, EERU and Clayton En-
                                                        vironmental Consultants, Inc. members for their contributions.
 84
AIR MONITORING

-------
        FIELD  SAMPLING FOR MONITORING,  MIGRATION
                      AND  DEFINING THE  AREAL EXTENT
                         OF  CHEMICAL CONTAMINATION

                                         JOHN M. THOMAS,  Ph.D.
                                                 J.R. SKALSKI
                                          L.L. EBERHARDT, Ph.D.
                                                M.A. SIMMONS
                                         Pacific Northwest Laboratory
                                             Richland, Washington
INTRODUCTION

  As part of two studies funded by the U.S. Nuclear Regulatory
Commission and the USEPA, the authors have investigated field
sampling strategies and compositing as a means of detecting spills
or migration at commercial low-level radioactive and chemical
waste disposal sites and bioassays for detecting contamination at
chemical waste sites.
  Compositing (pooling samples) for detection1 is discussed first,
followed by the development of a  statistical test to determine
whether any component of a composite exceeds a prescribed max-
imum acceptable level.2 Subsequently, the authors explore  the
question of optimal field sampling designs2 and present the fea-
tures of a microcomputer program designed to show the difficul-
ties in constructing efficient field designs and using compositing
schemes.3 Finally, they propose the use of bioassays as an adjunct
or replacement  for chemical analysis  as a means of detecting and
defining the areal extent of chemical migration.
COMPOSITING
  A working definition of a composite sample obtained from
commercial radioactive low-level or chemical waste sites (CLLCW)
might be the  mass of air, water, biota or soils resulting from pool-
ing several individual samples together before radioanalysis is
done. Compositing samples from CLLCW sites will become attrac-
tive when the cost of a single analysis is large  relative to costs of
collecting, pooling and adequately mixing samples.
  One purpose of compositing is to obtain an estimate of average
concentration of a radionuclide or chemical  in some biotic  or
abiotic component which might move off-site." In contrast,  an-
other purpose might be detection  of on-site  spills, areas of  un-
acceptably high radioactivity or possible radionuclide migration
during routine site operation.4  The desired outcome in  the first
case is an estimate of concentration and an appropriate  estimate
of variance. When compositing for detection, the desired outcome
is a statement of the probability that none of the samples making
up the composite contained more than some amount (e.g., 1 nCi/g
or ppm) and a confidence limit for the probability estimate.
  Several papers which deal with compositing for detection have
been previously reviewed.1'2 Some possible scenarios for detecting
spills at CLLCW sites are illustrated in Figure 1. All 16 soil samples
are composited  (mixed) in scenario 1, random selection of four
samples from the entire site make  up each composite depicted in
scenario 2, while in scenario 3 the compositing is done within four
selected site locations (A through D, possibly based on some prior
knowledge). In all these scenarios a subsample is used for analysis.
  The apparent spills or hot spots are shown as the open circles in
the figure. For purposes of illustration, it is assumed that when one
hot spot is composited with three background samples, the result-
ing contamination cannot be detected in the subsequent analysis.
  In scenario 1,  detection of spills would actually depend on  the
level of contamination, the dilution by 12 background soil samples,
the ability to mix 16 samples uniformly and the subsample size
(when aliquots are analyzed). Under the assumption above,  the
contamination in the scenario 1 sample would go undetected. A
spill probably would be detected under scenario 2 (unless only one
hot sample happened to be selected in each of the four composites),
but the location would have to be determined by an additional
analysis of the four individual components (perhaps only half of
each of the original four samples was used to form the composite).
Site B would be  identified as contaminated under scenario 3, but
Site C could be  missed. Many other possible scenarios could be
constructed. Thus, a generalized and statistically based strategy  for
compositing is evidently needed.
                                        Soil Samples

                                         • Background Concentration

                                         O High Concentration
    Composite All 16
     Soil Samples
Composite Random Sets ol 4
     Samples
Composite 4 Soil Samples for Each
  ol Sue Locations A. B. C & 0
   Spill is not Detected
                A Small Probability That at Least One
                Spill Will B« Dtltclld Sin Undefined
                Unless. Components Are Individually
                Anelyied
                    Sue B Contains at Least One Spill
                    Sample Spill in Sue C is not
                    Detected
                         Figure 1
 Hypothetical Example of Compositing to Define Spills at a Hypothetical
    Waste Site. To interpret the analytic results, assume that one soil
component (high concentration) mixed with three background soil samples
        is not detectable by the analytic technique employed.
                                                                                          AIR MONITORING
                                                      85

-------
DETECTING A MAXIMUM ACCEPTABLE LEVEL

  Compositing samples can result  in appreciable cost reduction
when chemical or radiochemical analyses are expensive.' However,
the loss of information due to grouping samples and the subse-
quent averaging of contaminant levels needs to be minimized (e.g.,
compositing samples can result in conta.-.l.iant dilution where it is
not detected).
  Excessive contaminant dilution can be minimized by specifying
a minimum detection limit for the analytic  procedure (MDL) and
a maximum acceptable level of contamination within the substrate
(MAL). Under these constraints, the maximum number of samples
that can be grouped into a single composite is:
         n     MAL
              MDL
                                                          (1)
 provided that perfect mixing can be assumed or large masses of
 material can be analyzed.
   By limiting the size of a composite to n components, a sample
 just at the MAL will still be above detection limits if mixed with
 (n - 1) samples with very low concentrations. The following three
 analytic outcomes  may  occur in  the  analysis of a  composited
 sample (Fig. 2):
 •Observed concentration is below detection limits so  we conclude
  all n components are below MAL.
 •Observed concentration is between MDL and MAL, indicating the
  possibility that one or more components may be above the MAL.
 •Observed  concentration is above MAL, indicating  that  one or
  more components are  contaminated  at a  concentration above
  MAL.
   In the first case  above, a null hypothesis  of  no contaminated
 samples can be accepted with probability 1, while in the last case a
 null hypothesis of no excessive contamination can be rejected with
 probability 1. However,  the intermediate condition (second case)
 requires further evaluation to determine whether there is a  high
 probability of components which contain excessive contamination.
   When the parameters  for the distribution  of background  con-
 centration (i.e., mean and variance) are known, a statistical  test
 for  group testing can be performed based on the observed com-
 posite concentration. This group testing procedure is  most applic-
 able when chemical analyses are expensive, frequency of contam-
 ination is low, contamination levels are well above background,
 uncertainties of field sampling are great and numerous samples
 are collected. The significance level of the group test can be inter-
 preted as the  probability of declaring that one or more of the n
 components in a composite exceeds the MAL when  in fact  they
 do not.  Composited samples with significance levels below  5%,
 or some other cutoff point, would be candidates for further lab-
 oratory analysis of individual components (assuming some of each
 was saved) to determine which components were actually contam-
 inated.

              Safety ol n Componems in a Composite
     All Are Sale
     Probability - 1
                On'.' or MUM.'
               M.iy Be Uns.lfu


              Compute Probability
                                              Oili' 01 MUM
                                               is Uns.ilr
Probability   1
                  MDL                MAL

          Analytic Concentration in the Composited Sample
                          Figure 2
    Concentrations of a Chemical or Radiochemical in a Composited
Sample Made Up of n Component Subsamples. The conclusion regarding
  safety depends on whether the composite level exceeds the minimum
 detection limit (MDL) or the maximum acceptable level (MAL) for the
                        contaminant.
                                                            A test statistic has been developed2 and preliminary tables (ob-
                                                          tained by computer simulation) have been prepared. Current work
                                                          is focused on either obtaining an analytic solution (or an adequate
                                                          approximation to the solution) of the equations which  would
                                                          allow a complete set of tables to be constructed. Currently, tables
                                                          depend on a five point entry consisting of the critical value for the
                                                          mean concentration of n-composited samples from a background
                                                          distribution with an estimated mean and variance, observed aver-
                                                          age  concentration for a particular composited sample, the  nth
                                                          order statistic for a sample of size n (i.e., the highest concentration
                                                          among the components in the composited sample),  maximum
                                                          acceptable level or concentration (i.e., possible values for MAL)
                                                          and the significance level of the test (type I error, often 5%).
                                                            The joint probability can be interpreted as the intersection of
                                                          events when a mean  composite concentration  equals or exceeds
                                                          the background critical level and all n individual components have
                                                          concentrations less than or equal to the maximum acceptable level.
                                                          Thus,  the probability  of falsely rejecting the null hypothesis  (all
                                                          components in the composite are below the MAL) can  be  found
                                                          using tables of the test statistic.
                                                            The results  of limited simulations show  that the test is highly
                                                          effective with large sample sizes (many components in a composite,
                                                          again  assuming perfect  mixing or the ability to  analyze large
                                                          samples)  where contaminant levels frequently exceed the MAL.
                                                          These  conditions of frequent and high contamination levels  are
                                                          likely to occur in regions of a commercial waste site where  the
                                                          prior knowledge (priors) for a spill or migration are highest.
 FIELD SAMPLING BASED ON PRIOR INFORMATION
   Historical records on land use or cursory surveys of waste sites
 often suggest possible locations where a search for spills can begin.
 This a priori information can improve the success rate of a survey
 scheme. However, in the absence of prior knowledge, more gen-
 eral strategies for sampling are needed.
   The most general approach  is systematic (uniform) sampling
 based  on  grid systems or lattice  designs.*  Recently, it has  been
 suggested  that the grid should be triangular for a  fixed area be-
 cause fewer samples may be required  (unless systematic dumping
 is suspected).' Nomograms for square, rectangular and triangular
 grids which can  be  used to select sample spacing for circular and
 elliptical spills have also been developed.1 In this study, the authors
 investigated how prior knowledge (prior(s) is standard terminology
 in Bayesian statistics] might be used to devise alternative sampling
 designs.
   A Bayes sampling approach can be defined as a sampling strategy
 chosen from a series of alternative schemes which best minimizes
 the risk of missing a spill. To illustrate the concept, competing
 sampling designs for a specific spill scenario were evaluated. The
 situation selected is  one where the location of a surface soil spill is
 known, but because of uncertainty about the mechanism by which
 the contaminant has spread, the extent of contamination over the
waste site  is unknown. Therefore, the initial objective of the sam-
pling program is to determine the size of the spill zone.
   Three prospective contaminated zones within the waste site are
identified  in Figure 3. Zone A, an area centered around  the sus-
pected spill site, has a high probability of encompassing part of the
spill. The second zone (zone B) is concentric about zone A and has
a slightly higher probability of containing the spill since it encom-
passes  zone A. This second area (zone B) is considered a margin
of safety in case zone A does not encompass the spill. The third
zone (C) is the remainder of the site and supposedly has very little
chance of contamination. The actual spill shown in Figure 3  indi-
cates that the initial suppositions were not entirely correct.
   The following three alternative  contamination  scenarios are
possible:
•Contamination is confined within zone A, 9% of the site area
•Part of the contamination has migrated beyond zone A but is con-
 fined within zone B, 8% of the site area and site A.
86
AIR MONITORING

-------
•Some contamination is in zones A,  B and  C (zone C contains
 83% of the site area).
  Only one scenario can exist  during the  sampling period. The
objective is to choose between the three scenarios.
        Waste Site
          Prior

          0.7

          0.2

          0.1
                            Figure 3
   Hypothetical Dispersion of a Contaminant Through Three Sampling
  Zones, A, B and C. The size and configuration of the zones are based on
    prior information and may be best guesses. In this case, the guesses
    were not entirely correct. The unlabeled decimal values are the areal
            proportion of the plume included in each zone.
   A probability has been arbitrarily assigned to each  scenario.
 Those probabilities, called priors, were arbitrarily  set at 0.70,
 0.20 and 0.10 (the probability that the  contamination is in zones
 A, B and C, respectively). Together, these probabilities are mutual-
 ly exclusive and their sum is one.
   Different sampling strategies at the hypothetical waste site were
 evaluated based  on  their  performance  under  each alternative
 scenario. Of the alternative sampling strategies possible, only those
 which allocate sampling efforts  based on areal zone size and/or
 the priors for the chances of contamination were considered. Using
 these two criteria, the following five possible strategies were de-
 fined:
 Sampling Strategy

 Si
 S2
 S3

 S4
 S5
Description

Allocation proportional to area
Allocation proportional to area and prior
Allocation proportional to area and in-
  versely proportional to prior
Allocation proportional to prior
Allocation inversely proportional to prior
   Using these basic data and a risk function (computed by sub-
 tracting the minimum loss among sampling strategies within scenar-
 ios, where loss is the probability of missing the contamination in
 any zone),2 each sampling  strategy was ranked for three possible
 scenarios: 30% of zone A  was contamined;  30% of zone A and
 15% of zone B were contaminated; and 30% of zone A, 15% of
 zone B and 5% of zone C were contaminated (Fig. 3).
   These scenarios correspond to one set of beliefs as to where and
 how far the spill could have migrated. The computed weighted risk
 functions in Table 1 indicate that the strategy with minimum risk
 results when  samples are allocated as  the  reciprocal of  the prior
 probability (85). Two other strategies (82, 84) had slightly higher
 weighted risk, while the strategy based on area/prior (83) resulted
                                            in the highest  risk.  Actual field  sampling allocation, given 100
                                            samples, under the five strategies would be very different (Table 2).
                                            None of the strategies would result in  an equal distribution of
                                            sampling effort over the site.
                                              Finally, the authors note that the work on compositing indi-
                                            cates  there should be more intensive sampling in suspected high
                                            concentration areas (high priors), while work on optimum designs
                                            yields the opposite result (that is, fewer samples in areas where
                                            priors for the spill are highest).  This indicates that an optimum
                                            strategy may exist and future research should be directed to obtain-
                                            ing such results.
                                                                        Table 1
                                                  Calculated Weighted Risk for Five Sampling Strategies and
                                                      Three Scenarios for the Example Spill in Figure 1
                                                         (See reference 2 for detailed calculations.)
                                                                                    Risk and Weighted Risk for Each Sampling Strategy (Sj)
Contamination
Scenario
30% Zone A
30% Zone A,
15% Zone B
30% Zone A, 15%
Zone B, 5% Zone C
Prior
0.7
0.2
0.1
Area
0.04
0.26
0.22
Area x Prior
0
0.16
0.14
S3
Area/Prior
0.70
0.79
0.75
Prior
0
0
0.52
S5
I/Prior
0.04
0.01
0
                                                                                    0.72
                                                                                              0.05    0.03
                                                                       Table 2
                                             Calculated sample allocations for five sampling strategies as a function
                                            of the percentage of the total site area represented by each zone and priors
                                               (0.7,0.2 and 0.1 for zones A, B and C, respectively). One hundred
                                              samples were allocated for each strategy. (See reference 2 for detailed
                                                                    calculations.)
                                                              Sample Allocation Based on Each Strategy
                                             Zone    Percent    Area    Area x Prior   Area/Prior
                                                                                            Prior
                                                                                                   I/Prior
A
B
C
9
8
83
9
8
83
39
10
51
1
5
94
70
20
10
9
30
61
                                                                            Sum
                                                                                     100
                                                                                            100
                                                                                                        100
                                                                                                                    100
                                                                                                                           100
DIGMAN—A MICROCOMPUTER PROGRAM

  DIGMAN3 was developed for site managers to illustrate the dif-
ficulties in sampling commercial radioactive low-level waste sites
and to allow cleanup  personnel to evaluate alternative sampling
strategies. This interactive program tests one's ability to locate a
contaminated area  and to determine its areal extent. The actual
field sampling design and the decision to use compositing are under
program user control.
  In the DIGMAN scenario, it is assumed that historical records
or a preliminary site survey indicate that contamination is present
and that the highest possible concentration is ten units per area. By
sampling the site soils, the extent of contamination must be  de-
termined and a decision made as to whether or not the contam-
inant has migrated off-site. Because of high laboratory costs, only
five samples (which can be composites) can be analyzed. Each of
the five permitted composite samples can be composed of from one
to nine component samples (e.g., up to nine samples can be com-
                                                                                                     AIR MONITORING
                                                                                                        87

-------
bined into one sample), but only the total sample may be analyzed.
Even though sampling is restricted to the waste site, the contam-
inated area  may extend beyond site boundaries. After sampling,
the site manager must determine the areal  extent of the contam-
ination. Since site  cleanup costs are assumed to be very high, the
smallest possible estimate of the contaminated area is desirable to
avoid condemning a larger area than necessary.
   Several different scenarios related  to sampling and costs can be
used. For instance, costs of collecting and analyzing  samples may
be high so the site manager would probably obtain the maximum
number of components per composite as well as all five composite
samples. In  another case, collection costs may be low but analyses
still expensive. Thus,  the site manager might take fewer than  five
composite samples,  analyze the  results and resample. The case
where collecting samples is expensive, while analyses  are relatively
cheap, would again result in selecting maximum numbers of com-
ponents and composites. When costs  of collection  and analyses
are inexpensive, compositing is not advantageous.
   DIGMAN also allows the site manager to resample after the first
sampling sequence is completed and the analytical results are avail-
able. This two or more stage  sampling is analogous to the  cir-
cumstances  where laboratory  turnaround is fast (and perhaps
cheaper than assumed in DIGMAN)  and  should  allow a  better
definition of the spill area.
   The waste site is simulated as a 40 x 40 grid. The  site manager
is given information that contamination exists at least at one point
on the waste site (this appears on the screen as  a darkened square
and is called a prior). However, the concentration  at this point is
not known  since in real  circumstances a site  manager will usually
only have sketchy information.
   The contaminated area is represented by an ellipse, because point
spills are generally moved by physical forces (e.g., wind or surface
water). The ellipse is generated using a bivariate normal distribu-
tion, and parameters affecting placement,  orientation  and  shape
are randomly determined.  Thus, the ellipse can be very small or
large  and of varying length and  width (e.g.,  very  "skinny" or
 "fat"). In addition, it is possible for part of the ellipse to be out-
side the defined waste site, an indication that contamination has
moved off-site.
   Sampling success is evaluated based on  the following three cri-
teria:
 •The proportion of the  contaminated  area (ellipse) within a user
  supplied circle. The  site manager supplies the radius and the esti-
  mated center of contamination.
 •The  fraction of the user chosen radius compared to the longest
  radius of  the ellipse (which indicates whether the manager  was
  conservative in estimating the size of the contaminated area).
 •The accuracy of the prediction that contamination was restricted
to the site.
 MAPPING CHEMICAL CONTAMINATION USING A
 PHYTOASSAY
    In order to demonstrate the usefulness of bioassays in chemical
 hazard assessment, a field study was conducted at  Rocky Moun-
 tain Arsenal (RMA) in Commerce City,  Colorado. Only the re-
 sults from photoassays of site  soils are presented here;  results of
 other bioassays are in the article by Thomas el al.ta
   The site  had  been used for  the manufacture of  anti-personnel
 gases, herbicides, insecticides  and  as an ordinance testing area.
 Over the years,  a myriad of organic and inorganic compounds
 were carried through ditches to a series of interconnecting holding
 basins for disposal. Thus, site soils  would be very expensive to
 completely  specify chemically and offered an excellent phytoassay
 opportunity.
 Study Site
   Four parallel transects were established near a waste trench, each
 beginning on the north bank of the trench and running south for
 approximately 90 m. A logarithmic sampling scale was used beyond
                                                         the south trench edge to locate sampling points, because contam-
                                                         ination probably moved by some physical means (e.g., wind or
                                                         water). The transects  were 15 m apart.  The first three sample
                                                         points of each transect fell within the trench and the fourth was on
                                                         the I Dp of the south bank. Sample numbers 5 through 9 were  15,
                                                         20, 30, 50 and 90 m, respectively, south of the north trench edge.

                                                         Soil Sampling
                                                           At most sampling points, a split spoon corer mounted on a hy-
                                                         draulic drill rig was used to take two soil cores, one from a depth of
                                                         0 to 15 cm, and a second from 15 to 30 cm. Each core was 7.6 cm in
                                                         diameter and together they weighed approximately 4 kg. Between
                                                         sampling points, the split spoon and drill bit were decontaminated
                                                         by  washing  with methanol  and rinsing with distilled  water. All
                                                         samples were placed in plastic bags,  sealed and labeled. The area
                                                         being sampled and any problems  encountered (e.g., mud, access-
                                                         ibility) dictated exactly how the cores were  taken and any varia-
                                                         tions on the basic sampling scheme.' The lettuce seed phytoassay
                                                         used is described by Thomas and Cline.''
                                                         Photoassay Results
                                                           Two seed mortality experiments (0-15  cm  and 15-30 cm  soil
                                                         samples respectively) were conducted and the maximum mortality
                                                         difference between  those soils  assayed  in both  experiments was
                                                         about 15%. Using this value as a cutoff point to assess mortality
                                                         differences at  the two depths showed (on inspection)  that lettuce
                                                         seed mortality differed (the 15-30 cm soil samples caused higher
                                                         lettuce seed  mortality). Moreover, seven samples exhibited much
                                                         higher mortality at the 15-30 cm depth compared to their 0-15 cm
                                                         fractions, suggesting  that the contaminants had  either migrated
                                                         below 15 cm or were purposely placed there.
                                                         Mapping Chemical Contamination
                                                           One way to depict lettuce seed mortality patterns at each depth
                                                         was to prepare  a contour map based on the observations. The
                                                         authors elected to use a relatively new statistical technique called
                                                         kriging developed for  use in the  mining industry and  used prin-
                                                         cipally in Europe and South Africa12-13 to calculate map contours.
                                                           Kriging is a weighted moving average technique that calculates
                                                         point estimates or block averages over a specified grid. The deriva-
                                                         tion of the kriging weights takes into account the proximity of an
                                                         observation to the point or area of interest, the structure of the ob-
                                                         servations (i.e., the  relationship of the squared difference between
                                                         pairs of observations and the intervening distance between them)
                                                         and any systematic trend or drift in the observations. Additionally,
                                                         kriging provides a variance estimate that  can be used to construct a
                                                         confidence interval  for the estimated observations. Contour maps
                                                         are prepared from the kriging estimates.
                                                           The results  of kriging  the 0-15 cm lettuce seed mortality are
                                                         shown in Figure 4. The  contamination  predicted by kriging was
                                                         greater at 15-30  cm compared to 0-15 cm, confirming the prelim-
                                                         inary analysis based on the observed data.
                                                           Maps similar to Figure 4 could be useful in site cleanup decisions
                                                         (especially when accompanied  by error  estimates). As a possible
                                                         scenario, if the 30^o mortality contour was  -elected as a criterion
                                                         for cleanup of the trench site, the area in Figure 4 enclosed by the
                                                         solid line would  be targeted for cleanup. Unfortunately, the clean-
                                                         up  decision would be different for the 0-15  cm and the 15-30 cm
                                                         soil fractions.  While this would complicate decision making, it is
                                                         the field situation that is complex, so decisions based on the 0-15
                                                         cm samples alone could have unwanted  consequences. In the cur-
                                                         rent case, samples below 30 cm would be needed to make a defen-
                                                         sible decision.
                                                          CONCLUSIONS

                                                            This paper summarized initial research on compositing, field de-
                                                          signs and site mapping oriented toward detecting spills and migra-
                                                          tion at commercial low-level radioactive or chemical waste sites.
 88
AIR MONITORING

-------
      90
      80
      70
      60
      50
% Mortality
D 0* -20

E3 20+-30
fl 30+ -50

H 50+-75
D 75+ -100
                                     Seed Mortality > 30%
                       15             30           45

                  Distance (m) From Northeast Ditch Corner
                           Figure 4
  Predicted Lettuce Seed Mortality Using Kriging. The observed data
        were obtained using 0-15 cm soil cores from the Rocky
                      Mountain Arsenal.
arithmic sampling sites (about 4000 m2) at two depths (0-15  cm
and  15-30 cm), mortality maps were obtained using kriging tech-
niques. Predicted mortality  (as well as observed  mortality) was
greater at the 15-30 cm depth. The results indicate  that phyto-
assay,  accompanied by kriging, can aid in site cleanup decisions
and  in defining the extent of contamination, particularly if  the
estimated contours are accompanied by kriging errors.

ACKNOWLEDGEMENTS
  The  technical  guidance of Dr. Edward O'Donnell (U.S. NRC)
and  Mr. William Miller (USEPA) is  gratefully acknowledged.
The  technical assistance of Drs. Dave Thorne and Bill Troutman
and  Mr. Brian Anderson  of the Rocky Mountain Arsenal was
essential to  completing portions of this work.  Jeanne Simpson
(Statistics Section, PNL) kriged the lettuce seed mortality data.
  This work was primarily sponsored by the U.S. Nuclear Regula-
tory Commission  under a Related Services Agreement with  the
U.S. Department of Energy Contract DE-AC06-76 RLO 1830 (TD
1634) and in part  by the U.S. Environmental Protection Agency
(TD 1598).  Although the research  described  in this presentation
was in  part funded by the U.S. Environmental Protection Agency,
it has  not been subjected to the Agency's required peer and  ad-
ministrative review and, therefore, does not necessarily reflect  the
views of the agency and no official endorsement should be  in-
ferred.
   Results  indicate that the significance  test developed to detect
 samples containing high levels of contamination when they are
 mixed with several other samples below detectable limits (com-
 posites) will be highly effective with large sample sizes when con-
 taminant levels frequently or greatly exceed a  maximum accep-
 table level. These conditions  of  frequent and high contaminant
 levels are most likely to occur in regies of a commercial waste site
 where the  priors (previous knowledge) about a spill or migration
 are highest. Conversely, initial investigations of Bayes sampling
 strategies suggest that  field sampling efforts should be inversely
 proportional to the priors (expressed as probabilities) for the occur-
 rence of contamination. In other  words,  fewer samples should be
 allocated at the probable source of a spill, with greater emphasis
 placed on  confirming the absence of contamination in suspected
 "clean" areas.
   Together, the joint results on group testing and sampling designs
 suggest that fewer but larger composite samples should be collected
 at the suspected source of a spill, while smaller and more frequent
 composite  samples should be taken in areas where uncertainty is
 greatest. By using this approach, the effectiveness of group testing
 is maintained along with the efficiency of the Bayes strategies in
 allocating field sampling effort. It is this prospect for coordinating
 the laboratory compositing and field sampling schemes which holds
 the greatest promise for efficient and cost-effective detection of
 spills and defining migration.
   The DIGMAN  microcomputer program was developed  to illus-
 trate the complexities in sampling waste  sites for spills or migra-
 tion. The  site-manager is  given prior knowledge that  a spill has
 occurred and that it has subsequently migrated through  or over the
 soil surface. In addition, the location is given for one point where
 some contamination is known to exist. Such an array of informa-
 tion may or may  not be available at actual sites.  The DIGMAN
 waste site provides 1600 possible sampling sites, clearly far fewer
 than would be available at an actual waste  site. Thus, the situa-
 tions depicted by DIGMAN are perhaps the simplest of the myriad
 of possible scenarios that might  be faced  by a site-manager. A
 floppy disc copy of the program can be obtained from the authors.
  In order to illustrate how phytoassay  results could  be  used to
map the toxic potential  of a chemical waste site, a small scale  field
study was conducted at the Rocky Mountain Arsenal. Based on the
results  from lettuce seed phytoassays of soil  samples from 36 log-
                    REFERENCES

                     1.  Eberhardt, L.L., and Thomas, J.M., Survey of Statistical and Sam-
                        pling Needs for Environmental Monitoring of Commercial Low-Level
                        Radioactive  Waste Disposal Facilities: A Progress Report in Response
                        to Task 1. PNL-4804, Pacific Northwest Laboratory, Richland, WA,
                        1983.
                     2.  Skalski, J.R., and Thomas, J.M., Improved Field Sampling Designs
                        and Compositing Schemes for Cost Effective Detection of Migra-
                        tion and Spills at Commercial Low-Level Radioactive or Chemical.
                        Waste Sites.  PNL-4935, Pacific Northwest Laboratory, Richland,
                        WA, 1984.
                     3.  Simmons, M.A., Skalski, J.R., Swannack, R. and Thomas, J.M.,
                        1984. DIGMAN: A Computer Program to Illustrate the Complexities
                        in Sampling Commercial Low-Level  Waste Sites for Radionuclide
                        Spills  or Migration.  NUREG/CR-3797, U.S. Nuclear Regulatory
                        Commission, Washington, D.C.
                     4.  Code of Federal Regulations, Part 61-Licensing Requirements  for
                        Land Disposal  of  Radioactive Wastes, 47 Fed.  Reg. 57463-57477
                        (Dec. 7, 1982).
                     5.  Schaeffer, D., Kerster, H.W. and Janardson, K.G.,  "Monitoring
                        Toxics by Group Testing." Environ. Management 6, 1982,467-469.
                     6.  Cochran, W.G.,  Sampling Techniques. 3rd  edition.  John Wiley and
                        Sons, New York, NY, 1977.
                     7.  Parkhurst, D.F., "Optimal Sampling Geometry for Hazardous Waste
                        Sites."Environ. Sci. Technol. 18, 1984, 521-523.
                     8.  Zirschky, J.  and Gilbert, R.O., "Detecting  Hot Spots at Hazardous
                        Waste Sites." Chem. Eng., 91, 1984.
                     9.  Thomas, J.M.,  Cline,  J.F.,  Gano, K.A., McShane, M.C., Rogers,
                        J.E., Rogers, L.E., Simpson, J.C. and Skalski, J.R., Field Eval.-
                        uation of Hazardous Waste Site Bioassessment Protocols. PNL-4614,
                        Vol. 2, Pacific Northwest Laboratory, Richland, WA, 1984.
                    10.  Thomas, J.M.,  Cline,  J.F.,  Skalski,  J.R.,  McShane, M.C., Simp-
                        son, J.C., Miller, W.E., Green, J.O., Callahan, C.A.  and Peterson,
                        S.A., "Characterization of Chemical Waste Site Contamination and
                        Its Extent Using Bioassays." (manuscript submitted),  1984.
                    11.  Thomas, J.M. and Cline, J.F., "Modification of the Neubauer Tech-
                        nique to Assess Toxicity of Hazardous Chemical.s in Soils." Environ-
                        Tox. and Chemistry (in press), 1984.
                    12.  Journal, A.G. and Huijbregts, C.H.H., Mining  Geostatistics. Aca-
                        demic Press,  New York, NY,  1978.
                    13.  Clark, I., Practical Geostatistics. Applied Science Publishers London
                        1982.
                                                                                                    AIR MONITORING
                                                                                 89

-------
                   SUBSAMPLING OF HAZARDOUS  WASTE

                                              R. SWAROOP, Ph.D.
                                                 O.S. GHUMAN
                                          Woodward-Clyde Consultants
                                              Santa Ana, California
INTRODUCTION

  The California Department of Health Services (DOHS) protocol'
for assessment of hazardous waste requires  that representative
samples be first analyzed for total concentrations of chemicals.
When testing for toxicity, if the 80% confidence limit of the total
concentration obtained from the sample analytical data exceeds the
total threshold limit  concentration (TTLC), then the  material
analyzed is considered hazardous  waste. If the 80% confidence
limit is below the soluble threshold limit concentration  (STLC),
then the material is not considered hazardous.
  However, when the 80% limit is  above the STLC and below the
TTLC, then a subsample must be analyzed using the California
Waste Extraction Test (CWET), and a new 80% confidence limit
must be calculated based on the results of the CWET analyses.  If
this new limit is higher  than STLC, then  the conclusion can be
made that  the waste is hazardous.
                         A AvontM
                                               MT-21
                                  CT-21        •
                                     •      MT-18
                          CB-12            MB-16
                                    MT-14
                                       •
                                            MB-11
                                  MT-OA    •
                                                MT-7
• Sample location*
                         Figure 1
 Soil Sampling Locations on the Site Investigated for Site Assessment
                                                       In this paper, the authors discuss the selection of subsamples for
                                                     CWET analyses as well as the statistical analysis of data for site
                                                     assessment in California. The data utilized in this paper are adapted
                                                     from an actual site assessment project, but the particular site and
                                                     other project details have not been identified because of a confiden-
                                                     tiality agreement.

                                                     SITE DESCRIPTION AND SAMPLING PROGRAM
                                                       The study site is a 40-acre tract in California (Figure 1) that has
                                                     been used for ranching, farming and oil production since 1920. A
                                                     few oil wells are still operating on the site. The current owner pur-
                                                     chased the tract in 1983 and planned development of the land. The
                                                     oily wastes at the site needed excavation to assess whether or not
                                                     they were hazardous.
                                                       Sample locations (Figure  1) were randomly selected in areas
                                                     where oil wells, storage tanks, sumps and/or pipelines were or had
                                                     been present. There was no visible surface evidence of these loca-
                                                     tions because the area was covered by vegetation.
                                                       The site assessment program consisted of the following  tasks:
                                                     review of available site data, interpretation of the historic aerial
                                                     photographs, preparation of  field program to collect soil samples
                                                     from  assigned  locations, chemical  analysis of  soil  samples and
                                                     statistical analysis of the data. The soil samples were analyzed for
                                                     several chemical constituents. For the purposes of this paper, the
                                                     authors limited their discussion to eight heavy metals: arsenic, cop-
                                                     per, chromium, nickel, lead, vanadium, zinc,  barium and cad-
                                                     mium.

                                                     CHEMICAL ANALYSIS DATA
                                                       Following the DOHS protocol, 24 soil samples were chemically
                                                     analyzed for the total concentrations of the eight heavy metals. The
                                                     chemical analysis results are shown in Table 1. The table shows the
                                                     sample locations, the  heavy  metal total concentration in the soil
                                                     samples  from these locations and the total and soluble threshold
                                                     limit concentrations (TTLC,  STLC)  for these metals as published
                                                     by  DOHS.' Since the authors' objective was to discuss the sub-
                                                     sampling, only barium  concentrations were selected for detailed
                                                     discussion. The subsampling  and the associated statistical analysis
                                                     procedure developed for barium was applicable to the chemical
                                                     concentrations of all the other heavy metals on the site.
                                                       The barium concentration in the 24 samples ranged from 34.6 to
                                                     538.0 ppm. In spite of such a large range of concentrations, most of
                                                     the values were close to 100 ppm. The frequency polygon and the
                                                     cumulative frequency polygon for these data are shown in Figure 2.
                                                     All the values are below the  barium TTLC value of 10,000 ppm,
90
AIR MONITORING

-------
                           Table 1
   Total Concentrations of Selected Heavy Metals in Waste Material
Sample
CB-5
CB-8
CB-12
CB-14
CB-1 5
CB-16
CB-17
CB-19
:B-2Q
CB-23
CB-21
•IT- 2
MT-3
17-4
MT-6
MT-7
MT-B
MT-9A
MT-9B
MT-14
HB-1 1
IB- 15
MT-18
MT-21
TTLC
STLC

40.2
30.2
36.2
41.7
24.6
34.1
29.2
26.4
26.0
29.3
16.9
42.8
4.6
2.1
24.0
23.7
44 . 1
3.2
22.7
14.7
2.2
35.9
3.5
19.4
500
5

7.6
13.1
7.9
15.0
10.4
16. a
6.7
7. 7
56.1
a. a
12.2
12.6
13.3
22.2
10.7
16.7
50.7
17.9
11.5
14.4
11.8
13.6
31.9
14.9
2.500
25
Heavy
17.2
20.3
16.9
22.6
16.6
31.9
15.5
15.0
13.5
25.2
14.6
23.7
21. a
22.5
15.2
19.9
27.3
47. a
15.8
14.5
14.3
19.3
21 .7
14.5
2.500
560
Metals
Ni
12.1
13.3
13.9
24.7
12.8
16.6
10.5
9.3
11.9
13.6
12. a
15.5
15.0
30.4
12.6
15.5
28.0
17.6
9.7
14.1
12.2
14.3
27.5
14.5
2,000
20
( mq/Jtg
Pb
5.2
4.8
5.6
82.5
10.3
6.9
4.8
5.7
10.7
4.1
45.8
5.0
0.5
0.76
5.3
14.8
49.3
2.2
14.8
28.4
<0.5
3.0
2.2
15.5
1,000
5
- ppm)
V
41.1
37.2
39.3
44.0
34.1
40.2
31.5
28.6
31. 1
30.3
29.0
47.4
47.5
39.6
32.1
36.0
47.7
31.7
32.5
27.2
31.9
39.3
49.9
29.9
2,400
24

Zn
45.8
43.9
43.1
62.6
45.6
39.2
36.0
53.1
90 .7
30.8
43.2
46.4
49.2
104.0
31.0
52.4
426.0
112.0
35.1
39.1
44.0
44.3
131.0
52.1
5,000
250

Ba C
58.8 <
76. 8 <
67.5 <
131.0
74.5
86.6 <
34.6 <
37.4
53.6 <
63.1
76.6
113.0
93.8
102.0
95.5
102.0
114.0 •
148.0
68.5
87.8
80.3
195.0
538.0
307.0
10,000
100

d
0.05
0.05
0.05
0.19
0.11
0.05
0.05
0.05
0.05
0.09
0.50
0.08
0.03
0.02
0.10
0.09
10.05
0.04
0.03
0.10
0.10
<0.05
0.11
0.02
10
1
 but there are a few sample values which exceed the barium STLC
 value of 100 ppm. Thus, the 80% confidence level of barium con-
 centrations is likely to be between 100 and 10,000 ppm.
   By the DOHS protocol, CWET test results  for soluble barium
 concentration in the soils were required to  assess if the material
 constituted a hazardous waste. Thus, a subsampling of sample
 population was needed to make such an assessment.
          FREQUENCY POLYGON OF Ba CONCENTRATION
   FHEO
(f of 8»mpl««)
              10O       ZOO       3OO        400
             X - Total Barium Concentration (ppm)
                                                       800
     CUMULATIVE FREQUENCY POLYGON OF Ba CONCENTRATION
              1OO       JOO       30O       400
             X - Total Barium Concentration (ppm)

                            Figure 2
             Observed Polygons of Barium Concentration
                                                       6OO
STATISTICAL RESULTS USED IN SUBSAMPLING

  The first statistical analysis was performed on the distribution of
the total barium concentration. The data shown in Table 1  and
plotted in Figure 2 indicate that the distribution is highly peaked
around 100 ppm, skewed and long-tailed toward higher concentra-
tion values. Such a distribution can hardly be assumed to represent
a symmetric normal distribution.2 Therefore, the mean value of
116.9 ppm and the standard deviation of 106.0 ppm shown in Table
2 could  not be satisfactorily used to calculate upper 80%. It  was
therefore necessary to evaluate a data transformation which could
approximately normalize the  barium data distribution and help
calculate an  80%  confidence value for  the hazardous waste
classification.
  Three  transformations  were  evaluated.   Assuming  that  T
represented total barium concentration in samples, the transforma-
tions were as  follows:
                                                                          X = Sin-'  x/T/106
                                                                          Y = ln(T)
                                                                          Z = -JT
                                                          (D
                                                          (2)
                                                          (3)
                                                                                            Table 2
                                                                  Transformed Values of Total Barium Concentration in Waste Materials
Samp) e
Locution
CB-5
CB-B
CB-12
CT-14
CT-15
CT-16
CB-17
CB-19
CB-20
CT-23
CT-21
MT-2
MT-3
HT-4
MT-6
MT-7
MT-8
MT-9A
MT-9B
MT-14
MB-11
MB-15
MT-18
MT-21
Mean
Standard
Deviat ion
T-Tolnl Harlum
Concentration X • Sin" % T/10 Y - In(T)
(ppm)
58.
76.
67.
131 .
74.
86.
34.
37.
53.
63.
76.
113.
93.
102.
95.
102.
114.
148.
68.
87.
eo.
195.
538.
307.
116.

106.0
a
8
5
0
5
6
6
4
6
1
6
0
a
0
5
0
0
0
5
8
3
0
0
0
1


0.4394
0.5021
0.4707
0.6558
0.4945
0.5332
0.3370
0.3504
0.4195
0.4551
0.5015
0.6091
0.5549
0.5787
0.5599
0.5787
0.6118
0.6971
0.4742
0.5369
0.5134
.0741
.3412
.2121
.8752
.3108
.4613
.5439
.6217
.9815
.1447
.3386
.7274
.5412
.6250
.5591
.6250
.7362
.9972
.2268
.4751
.3858
0.8001 5.2730
0.3291 6.2879
0.0040 5.7268
U.5836 4.5455

0.2122 0.6008
7 - s T
7.6681
8. 7636
B.215H
11.4455
B. 6313
9.3059
5.B822
6.1 156
7.3212
7.9436
8.7521
10.6301
9.6850
10.0995
9.7724
10.0995
10.6771
12.1655
8.2765
9. 3702
8.9610
13.9642
23. 1948
17.5214
20.2274

3.6900
   X is appropriate if T is binomial, Y if T is lognormal and Z if T is
 a  poisson variable.3 The transformed total barium concentration
 values, their means and standard deviations are shown in Table 2,
 To evaluate which of these transformations was suitable, all of the
 data shown in Table 2  were standardized by subtracting the mean
 from each value and dividing the difference by the standard devia-
 tion. These standardized values (Tj, Xj and Zj) are shown in Table 3.
   The observed distribution of each of the standardized variate
 against standard normal distribution (mean =  0, standard devia-
 tion = 1) was tested by the chi-square goodness of fit test.3 The chi-
 square values and associated results are graphically shown in Figure
 3. The probability of the chi-square exceeding 12.75 at 6 degrees of
 freedom is less than 0.0472; among the three distributions, Y =
 ln(t) was considered closest to the normal distribution. Thus, Y =
 ln(t) values were used  in the decision-making calculations.
   The second statistical analysis concerned the proportion of the
 samples whose observed T value exceeded the barium STLC value
 of 100 ppm. In terms of Y = ln(t), it referred to the number of
 samples whose values exceeded In (100) =  4.6052. The samples ex-
 ceeding this threshold  value were included in Group 1, and those
 not exceeding the threshold value were included in Group 2. These
 data with their means and standard deviations are shown in Table 4.
                                                                                                  AIR MONITORING
                                                                                                                            91

-------
                          Table 3
Standardized Values T,, X, and Z, of Barium Concentrations in Waste
    Materials Obtained by Subtracting the Mean and Dividing by
               the Standard Deviation of Table 2
                                                                                  Table 4
                                                            Total Barium Concentration T and Y = ln(T) of Samples and
                                                                          Their Group Classification
   Sampla
  Location
CB-15
CB-e
CB-12
CT-14
CT-15
CT-16
CB-17
CB-19
CB-20
CT-JJ
CT-21
MT-2
MT-]
MT-4
MT-6
MT-7
MT-6
MT-9»
MT-9B
MT-14
MB-H
MB- 15
MT-18
MT-21
-0.
-0.
-0,
0.
-0.
-0.
-0.
5481
3781
.4660
,1110
,4000
.2658
,7764
-0.7500
-0.
-0.
-0.
-0
-0
,5972
,5075
.1802
.0168
.2179
-0.1406
-0
-0
-0
.2019
.1406
.0274
0.2914
-0
-0
-0
0
3
1
.4566
.2745
.1451
.7168
.9726
.7914
-0.
-0.
-0,
0,
-0,
-0,
-0.
-0.
-0,
-0.
-0.
0
-0
6795
1841
,5120
,1402
4195
,2375
.1621
,0990
.7710
.6056
.1869
.1202
.1152
-0.0211
-0
-0,
0
.1117
.0211
.1329
0.5149
-0
-0
-0
0
0
0
.5156
.2201
.1108
.0201
.5113
.9811
-0.
-0.
-0,
0,
-0.
-0,
-1 ,
-1 ,
-0,
-0,
-0,
0
-0
0
0.
0
0
0
-0
-0
-0
1
2
1
7846
1400
,5549
,5488
.1906
1401
6671
5176
,»1B7
,6671
,1444
.1028
.0072
.1121
.0226
.1121
.1174
.7518
.5105
.1172
.2658
.2109
.9001
.9662
-0
-0
-0
0
-0
-0
-1
-1
-0
-0
-0
0
-0
-0
-0
-0
0
0
-0
-0
-0
1
1
1
.6916
.1967
.5451
.1101
.4125
. J497
.1776
.1141
.7876
.6189
.1998
.1091
.1470
.0147
.1211
.0147
.1219
.5252
.5287
.2)21
.1412
.0127
.5142
.9767
          -1.OC74 -0 5«S« -0.1787 ° 01707  0 5«5«  1.0674
                          Figure 3
        Distribution Functions of the Standardized Variables
  These two statistical results, (1) ln(T) was approximately nor-
mally distributed and (2) a fixed STLC value partitioned the sample
population into two  mutually  exclusive Groups  1 and 2,  were
utilized in the subsampling and in the site assessment.
SUBSAMPLING AND THE DECISION PROCEDURE

  The subsampling procedure  has  been formalized in symbolic
terms. The symbols have then been replaced by numbers using the
statistical results of the previous section to make an assessment of
material characteristics.
  Suppose n representative samples are obtained from a site. Let
TI, T2	Tn be their total Barium concentrations. It  is then
assumed that Yj = ln(T,)	, Yn  = ln(Tn) are random samples
from a normal distribution characterized  by a mean and a standard
deviation. The best_statistical estimates of the mean and the stan-
dard deviation are Y and Sy computed from the formulae:
s«n)il •
| .oca IJ.O"
CB-5
CII-8
CF1-12
CT-14
CT-15
CT-16
CD- I 7
CB-19
cn-20
CT-21
CT-21
MT-;
MT-J
HT-4
MT-6
MT-7
MT-8
MT-9A
MT-9B
HT- 14
HR-1 1
MB- 1 *>
HT-ia
MT-2 1
Number ol
M«an
Standard
r-Totfil IIMrtum
' .MM «nt t AI Ion
Ijipl") Y • InIT)
58
76
67
1 II
74
86
14
17
53
61
76
II 1
91
103.
95.
102.
114.
140.
68.
87.
80.
195.
518.
)07.
Sampla*

8 .0741
8
5
0




I
6
o
8
Q
}
0
0
0
5
8
)
0
0
0
.1412
.3121
.8753
.1108
.461 )
.1439
.6317
.9815
.1447
.1186
7374
.5413
. 675O
.5591
.62)0
.7J62
.9973
.2268
.4751
.1858
.2710
.2»79
. 7268
M
4.545i
Devlat Ion O.600A
(V • 4. 4052 ) ar
(T . IOOISTI.cn
Group 1



* S75J


--
__
._
--
4.7274
--
4 . <. i;o
..
4.6250
4.7)62
4.9972
--
--
--
5.27)0
».2»79
5.7268
9
i.CH'l
0.4720
r y« 4 -«iO'i2 ) nt
(T , IDnlSTIS-ll
Croup 3
4.0741
4.1412
4.2131
--
4.1108
4-4611
1.5419
1.4217
1.««I5
4.1447
4 )i«6
--
».54I3
--
4.4591
--
--
--
4.2268
4.4751
4.1858
--
--
"
IS
4.2145
0.1051
Y = E Y/N
y2 = E(Yi-Y)2/(n-l)
                                                         (4)
                                                         (5)
  From Y~and Sy values, the estimate of the average T = T and the
upper 80% T value = T go are obtained by the following relations:

        T = exp [Y]                                      (6)

        T.80 = [Y + tS/VnJ                            (*>
where t is the 80% 't'  value with (n- 1) degrees of freedom ob-
tained from the statistical tables. Then, the conclusion that waste is
hazardous or non-hazardous  is made by the following two-step
decision procedure.
Step 1.
    A
  If T go is greater than TTLC, then waste is hazardous.
    A"
  If T go is smaller than STLC, then waste is  non-hazardous.

  If T go is in between STLC and TTLC, go to Step 2.
Step 2.

  Perform CWET analysis for extractable barium in m samples out
  of a total  of n samples (m is less than or equal to n).  Let W,,
  W2,  ...  Wm be  the  extractable barium concentration values.
 .Then compute appropriately an upper 80% confidence
W gQ. If W go is greater than STLC, the waste is hazardous. If'
  is less or equal to STLC,  then the waste is  not hazardous.

  The procedure to select m samples out of a  total of n samples at
Step 2 addresses the subsampling problem. A  subsampling and the
related decision  procedure  is performed under the following
guidelines.
  a) Only m samples from the total n are to be selected for CWET
analysis.
  b) The proportion -p" of Tj values exceed STLC. The r samples
whose Ti exceed  STLC belong to Group 1, and r = np . The re-
maining (n - r) samples with T| less or equal to STLC belong to
Group 2, and n - r =  n  (1 - p).
  c) Select randomly mp samples from Group 1  and m  (1 - p)
from Group 2. Identify these m samples.
  d) Perform CWET analysis and assume the extractable barium
concentrations are Wj, Wj, ... Wm.
  e) For these i  = 1, 2, ... m samples, make transformations Vj =
 In (Wj) and  relable Y|  = In (Tj) as Uf. Calculate the two means 0,
 92
 AIR MONITORING

-------
 V, the two standard deviations S^, Sv and one correlation coeffi-
 cient r between Uj and V;.
   f) Use the five statistics described in e above and the two  Y
 statistics (Y, Sy) to calculate W and W80. If W 80 exceeds STLC,
 then it can be concluded that the waste is hazardous; otherwise, the
 waste is not hazardous.     A
   The calculation of W and W 80 is performed by the following for-
 mulae.4
W = exp [ V + r (SV/SU) (Y - V)] = exp [V]

    = [SV2(1 - r)/m] [l-(n  -
  .so = exp [V + t Sy]
                                                            (8)
                                                            (9)

                                                          (10)
where t is the 80% 't' value at (m - 3) degrees of freedom obtained
from the statistical tables.
   The decision procedure described above  has three important
features. First, it utilizes the correlation coefficient between total
and extractable barium concentrations. Second, it also accounts for
the  difference between  the  average of all the  samples  and the
average of subsamples. Finally, it further balances the subsampling
between the Groups 1 and 2 in relation to the population propor-
tions of samples which exceed or do not exceed particular STLC
value.
   Since  m  subsamples are obtained by randomization, this pro-
cedure thus ensures that the estimates calculated from the data re-
main statistically unbiased. In contrast, a selection of all m sub-
samples from Group 1 alone or any subjective method is likely to
add either  a positive or  a negative bias  for hazardous  waste
classification.

CONCLUSIONS FOR THE SITE
FROM SUBSAMPLING

   The subsampling and the decision procedure presented  sym-
bolically in the previous section were used to make conclusions
about the presence of hazardous waste on the site. There were n =
24 samples  with locations as shown  in Figure 1 . The total barium
concentrations (T) at these locations are given in Table 2. Project
considerations stipulated that at most m =  10 samples could be
subjected to CWET analysis. As shown in Table 4, there were  9
samples  in Group 1  and 15 in Group 2. Thus, proportion 'p' was
9/24 =  .375.
   A random selection of 10 (.375) ^ 4 samples from Group 1 and
10 - 4  =  6 samples from Group 2  was made to perform CWET
analysis. These 10 random sample locations and the total (T) and
soluble (W) concentrations of barium in the soil samples at these
locations are shown  in Table 5. Table 5 also contains data on the
means, the  standard deviations and the  correlation between the
transformed U and V values.
  The positive correlation  coefficient of  0.9362 indicated a very
strong relationship between the total and soluble barium concentra-
tions. This  demonstrated that at  this particular site the soluble
barium concentration in the waste depended only on the total con-
centration and on no other factors related to waste constituents.
                                                                                     Table 5
                                                           Total and Soluble Barium Concentrations (ppm) in the 10 Subsamples
                                                                 Selected from a Total of 24 Samples of Waste Materials
                                                             S a mp1«
                                                            Location
CB-e
CT-15
CB-17
CB-19
HT-6
KT-14
HT-8
MB-15
MT-18
MT-21
Number of
2
2
2
2
2
2
1
1
1
1
Subiamp
76
74.
34.
37.
95.
87.
114.
195.
538.
307.
l«s Group
.8
.5
.6
i 4
,5
8
0
0
0
0
1
                                                                                          Soluble W
 42
 29
 16
 21
 62
 36
 56
153
162
 97
4.3412
4.3106
4.5439
3.6217
4.5591
4.4751
4.7362
5.2730
6.2B79
5.7268
3.7377
3.3673
2.7726
3.0445
4.1271
3.5835
4.0254
5.0304
5.0876
4.5747
                                                            Number of Subaamplda Group 2 -  4

                                                            Total Number of Subaamplea   -  10
                                                                                                  0 • 4.6876   V - 3.9351

                                                                                                  Su- 0.8663 Sv- 0.7896
                                                                                                  Correlation between (U,V)
                                                                                                        - 0.9362
                                                            The decision to classify the waste as hazardous waste or not was
                                                          made by substituting the following values in the formulae for W
                                                          A
                                                          W go of the previous section.
                                                                     n = 24
                                                                     Y = 4.5445
                                                                     V = 3.9351
                                                                     Sv = 0.76896
 r = 0.9362
 m =  10
 U = 4.6876
 Su =  0.8663
                                                          Substituting the  above data in appropriate  equations  yields the
                                                          following results:

                                                          W = exp [3.8138] = exp [V]  = 43.3 ppm
                                                          Sy2 = (0.1764)2
                                                          W.80 = exp [3.8138  + 1.154(0.1764)] =  55.6 ppm

                                                            Since the STLC (100 ppm) for barium is much larger than the
                                                          55.6 ppm value calculated above, it was decided that the waste was
                                                          not hazardous due to barium concentrations. Similar subsampling
                                                          and calculations  were performed for other heavy metals. On the
                                                          basis of the  results, it  was similarly concluded that  the  waste
                                                          material was  not hazardous for all  contaminants analyzed given
                                                          California DOHS toxicity considerations.

                                                          REFERENCES

                                                          1. California Assessment Manual for Hazardous Wastes (Draft). Septem-
                                                            ber 28,1983, p. 103-104.
                                                          2. Johnson,  N.L. and Kotz,  S., Continuous  Univariate Distributions-1,
                                                            Houghton Mifflin Company, Boston, MA,  1970.
                                                          3. Snedecor, G.W. and  Cochran, W.G., Statistical Methods, The Iowa
                                                            State University Press, Ames, Iowa, 6th edition, 1973.
                                                          4. Cochran,  W.G., Sampling Techniques, John Wiley & Sons, Inc., New
                                                            York, NY, 2nd edition, 1966.
                                                                                                    AIR MONITORING
                                                                                                                              93

-------
                         QUALITY  ASSURANCE AUDITS  OF
                             FIELD  SAMPLING  ACTIVITIES

                                                  K.W. BROWN
                                    U.S.  Environmental Protection Agency
                                                Las Vegas, Nevada
                                                   D.S. EARTH
                                               University of  Nevada
                                                Las Vegas, Nevada
                                                   B.J. MASON
                                                     ETHURA
                                               Derwood, Maryland
INTRODUCTION

  Data from monitoring and sampling programs cannot be eval-
uated and interpreted with confidence unless adequate quality
assurance (QA) methods and procedures have been incorporated
into the program design. Adequate QA requires identification and
quantification of all sources of error associated with each step of
the monitoring and sampling effort. The identified sources of error
can then be analyzed using appropriate statistical tests yielding
estimates of the various components of variance.
  To date the  most highly developed aspects  of QA undertaken
in support of monitoring and sampling programs deal with analyti-
cal  procedures. Due to the complexity of designing adequate en-
vironmental monitoring programs, i.e., identification of a con-
taminant distribution in a heterogeneous environment such as the
soil system, the QA applied to the analytical procedures, even
though necessary, is not sufficient in itself to assess variability with-
in the sampled  environment. The analytical error may account for
only a small portion of the total variance. It is clear that a compre-
hensive QA program is required for the sampling portion of a mon-
itoring effort.
  In 1979 the USER A initiated a policy that required all USEPA
laboratories, program offices and regional  offices to prepare QA
program plans  for  all monitoring and measurement activities that
generate and process environmentally related data for agency use.
In 1980 the USEPA's Office of Monitoring Systems and Quality
Assurance (OMSQA)  issued guidelines identifying specifications
for  preparing  QA project plans.'  These guidelines identify and
describe 16 essential  elements  that  all QA project plans  must
address, the format to be followed for addressing these  elements
and how the plans would be reviewed and approved.
  One of the essential guideline elements identified is the require-
ment to conduct program audits. The following sampling and mon-
itoring activities conducted by USEPA are  examples of programs
requiring high quality, decision making data. As such, they are re-
quired to have a comprehensive audit program.
•Emergency cleanup operations
•Remedial response operations
•Preliminary data collection investigations
•Enforcement data collection investigations
•Regulatory purposes
•Research and technology transfer studies
  The principal function of the  overall QA program is to  assure
that proper design techniques are being implemented and that ade-
quate QA measures are being employed so  that the resulting data
                                                     will be of acceptable quantity and quality to satisfy program re-
                                                     quirements.' The QA audit function is not intended to evaluate the
                                                     technical merit or to verify the scientific validity of the sampling de-
                                                     sign, sampling devices or program protocols. Its purpose is to en-
                                                     sure that the methods and procedures identified in the programs,
                                                     protocols and QA plan are in place and are being followed.
                                                       The QA audit function is not  intended to threaten, intimidate
                                                     or abuse sampling/monitoring personnel in the performance of
                                                     their duties.  It should,  howe\er, verify that specified operating
                                                     procedures are implemented and maintained throughout the dura-
                                                     tion of the sampling program.
                                                       In this paper, the authors discuss those aspects  dealing  with
                                                     program audits relating  only to the sampling portion of the  total
                                                     monitoring program.
                                                     PURPOSE

                                                       As previously stated, the purpose of an audit is to ensure that the
                                                     protocols identified in the Project Plan and QA Project Plan are in
                                                     place and functioning well.
                                                       Specifically, the audit should:
                                                     •Verify that the sampling methodology and QA measures are being
                                                      performed in accordance with program requirements
                                                     •Verify that project documentation is in order, i.e., records, chain-
                                                      of-custody forms, analytical tags, log books, work sheets, etc.
                                                     •Verify the availability and presence of key project personnel and
                                                      their qualifications
                                                     •Identify QA problems
                                                     •Recommend corrective actions, if necessary
                                                     • Follow-up on previous recommendations
                                                     •Provide a written report of the audit
                                                       An audit should normally be designed, announced in advance
                                                     and planned with  the Project Officer of the sampling program
                                                     rather than being a surprise inspection. The reasons for this are:
                                                     •A surprise inspection may cause confusion among the field per-
                                                      sonnel and the other Agency observers
                                                     •A surprise inspection may hinder the field operations
                                                     •Due to the hazardous nature of many field projects, unannounced
                                                      visits could increase the risk of accidents
                                                     •Key personnel and/or log books and sampling records may not be
                                                      available
                                                       An unannounced audit would be advantageous only  because of
                                                     the element of surprise. Thus, there would be little time to correct
                                                     any  problems or deficiencies  occurring.  Unannounced  audits
94
AIR MONITORING

-------
should be performed only if there is information indicating that
there are serious problems with the sampling program.

AUDIT TEAM

  The audit team should have at least two people. The size of the
team will, of course, depend upon the extent of the operations be-
ing audited. However,  any operations which are potentially haz-
ardous should be performed using the buddy system.
  As  the  disciplines required to perform sampling  activities are
quite  varied, so too must  the audit team have a variety of tech-
nical expertise.
  Technical backgrounds required may include the earth sciences,
chemistry, engineering, health and  safety, biology and environ-
mental science. As far  as possible, the team should be composed
of specialists having overlapping experience in various  fields of
science and engineering related to the project to be  audited. The
team  also must consist  of mature professionals. The process of re-
viewing other  people's work and making constructive, objective
evaluations requires that the team members have both character-
istics. The additional aspect of a hazardous environment requires
that personnel be alert, safety conscious and possess a high de-
gree of professionalism.3
   The audit team should report to an Audit Program  Manager
who has overall responsibility for the audit and review of the final
report. The audit team leader,  selected by the audit program man-
ager, is primarily responsible for leading the team through prepa-
ration, the site visit and the preliminary report preparation.
 AUDIT PROGRAM MANAGER

   The Audit Program Manager should possess technical as well
 as managerial talents. Since the team must consist of professionals
 with a variety of scientific and engineering backgrounds, the back-
 ground of the manager should be as multidisciplinary as possible,
 preferably centered around environmental science. At a minimum,
 it is recommended that the Audit Program Manager have a Bach-
 elor's  degree in a scientific or engineering field or equivalent re-
 lated professional experience,  three years of experience as an au-
 ditor and one year of experience as an audit team leader.3
   The Program Manager selects team members and a Team Lead-
 er, makes assignments to individual members, assists the Leader
 where  necessary in preparing for the audit and approves all plans
 and reports. He is additionally responsible for retaining all records
 and reports of the audit proceedings.
 AUDIT TEAM LEADER

   The Team Leader should be selected from team members who
 have participated in a number of audits and have demonstrated
 clear managerial and leadership  qualities. At a minimum, it is
 recommended that the Team Leader have a Bachelor's  degree
 or three to eight years of relevant work experience in a scientific
 or engineering field and two years of experience as an audit team
 member. The Team Leader receives his assignment from the audit
 Program Manager, helps select team members, makes assignments
 and leads the team  in preparing, conducting and reporting the
 results of the audit.
AUDIT TEAM MEMBERS

   Each team member should have a Bachelors' degree in an appro-
priate scientific or engineering discipline. In addition,  each mem-
ber should have at least one year of experience in performing field
sampling. The team members work with the Leader in preparing
for the audit, conducting the audit and reporting the results of the
audit.
TRAINING

  Audit team training should be similar to that required for other
personnel involved in  hazardous waste site/facility investigations.
The following subject areas should be included in the personnel
training program.3
•Performing an audit
•RCRA/CERCLA Regulations including the rights of inspectors
 and owner/operators of hazardous waste sites/facilities
•Safety protocols including removal, decontamination and disposal
 of clothing and equipment used during site visits and  use and re-
 strictions of clean areas
•Safety equipment including the use of  respirators and self-con-
 tained breathing apparatus and protective clothing
•First Aid/Cardiopulmonary resuscitation
•Site-specific contingency and evacuation plans
•Legal ramifications of the audit including requirements of chain-
 of-custody, preservation of evidence and witness and testimony
 responsibilities4
•Risk assessment, recognition and evaluation of extent of hazards,
 methods used to  control risks and chemical compatibilities/re-
 actions
•Personal hygiene including prohibitions against eating, drinking
 and smoking and the effect of facial hair on respirators5
•Certification  at the intermediate and/or advanced level  of the
 USEPA's Health and Safety Training Program6
IN-HOUSE AUDIT PREPARATION

  The audit team should prepare to conduct the audit by review-
ing project documents including QA plan, protocols and progress
reports. In reviewing the documents and preparing for the audit,
checklists should be prepared which will aid the audit team in iden-
tifying  procedures in the field which  are crucial  to  the project
goals. Preparations for the field audit/site visit should  also include
a review of health and  safety requirements and field equipment
needed for the audit. Final preparations should include commun-
ications  with the  Project  Officer regarding the  anticipated
schedule, activities to be observed, any current problems and assis-
tance with health and safety aspects including the availability of
on-site safety equipment for the audit team.
  Specific documents that should be examined include:
•Project plan
•QA Project plan and QA reports
•Protocols and methods
•Previous audit reports from other offices or agencies
•Project and progress reports
•Contract and proposals
•Documents to provide  background information on the site (e.g.,
 RCRA  permit   applications,  preliminary  assessment reports,
 groundwater monitoring plans,  maps, photographs, etc.)
•Health  and safety plan including contingency and evacuation
 plans
•Chain-of-custody procedures and documents
  For projects involving RCRA-regulated sites, background docu-
ments should include the facilities Part A and applicable sections
of the Part B permit (e.g., Waste Analysis Plan, Groundwater
Monitoring Plan, etc.)  applications.  If the project involves  a
CERCLA site, a preliminary site assessment report or other infor-
mation may be available.
  The  project documents should be reviewed to  understand the
overall project goals so that activities which are critical to those
goals may be audited. Assignments for the team members should
be based upon the site  activities to be audited and the available
team members' backgrounds. Assignments for the audit in-house
preparation and  field activities should be matched  to  the team
members whose  experiences best  suit  these assignments. For ex-
ample,  some assignments may  require more experience and ex-
pertise in soils than in the engineering or geology disciplines.
                                                                                                  AIR MONITORING
                                                                                                                            95

-------
  The products of the in-house audit preparation should be  the
following:'
•Assignments for the  team members during both the preparation
 and field audit phases
•Checklists to identify and verify performance of critical steps in
 activities
•List of equipment and supplies needed during the audit
•Schedule of activities for the site visit including the introductory
 meeting with senior field personnel, the various audit activities, a
 session for the team to prepare for the debriefing and  the  de-
 briefing of site personnel
CONDUCTING THE AUDIT

   Arrangements should be made between the Team Leader and the
Project Officer for the site visit to conduct the audit. Prior to the
site visit, a schedule  of the audit which will not interfere with the
project should be arranged. A list of recommended procedures to
be followed when conducting the on-site sampling audit is found
below.5
 Do:
 •Do, upon arrival at site, immediately identify audit team per-
  sonnel to the Project Officer or most senior project person on
  site
 •Do meet with the project personnel and review the intended work
  schedule identifying which on-site personnel and operations will
  be involved in the audit
 •Do review all safety requirements, hazards and safety equipment
  which will be used on-site
 •Do conduct  the audit during normal working hours and at  the
  convenience of the owner or manager of the site and the Project
  Officer
 Do Not:
 •Do not discuss judgments with site personnel
 •Do not participate during the visit; the audit team members are
  strictly observers
 •Do not hinder operations
   After meeting with the on-site  Project Officer  and reviewing
 the audit schedule and tasks, each  team member should start per-
 forming his audit functions using the checklists.
   Whenever  possible  inventory the  sample bank records and
 archived samples to verify that the documentation is in order and
 sufficient to establish the disposition of any  sample collected.
 Trace the flow of specific samples through the system. Records to
 review include: Chain-of-custody (COC) forms. Sample Tags, Cus-
 tody Seals, Shipment Forms, Logbooks  and Archived Samples.
 Logs must be clear and concise.  Logbooks changes made by field
 personnel should be initiated and lined through so that the orig-
 inal entry is  still visible.  Problems should be documented in  the
 logs.
   Verify personnel identified in  the  Project Plan,  QA Plan and
 contract proposal.  Include all managers, middle managers,  pro-
 fessional specialists and first line field supervisors.
   Observe activities carried out by the sample bank custodian(s).
 Before  accepting custody of any samples, sample bank personnel
 should check  to make sure that:
 •Each sample has a completed sample collection tag attached
 •Each sample is identified on the COC form
 •A sample/site description form or record has been completed for
  each sample
 •Discrepancies are corrected
   Observe sampling and sample handling procedures first-hand.
 Sample handling procedures may include  drying, sieving, mixing,
 compositing,  splitting, packaging and shipping.
  Observe housekeeping: safety, decontamination, accident docu-
mentation and security.
  Observe sampling equipment  and containers and  the cleaning
and storage of sampling equipment.
  Observe the collection procedures preparation and frequency of
collection of field blanks, replicates, splits and spikes if any.
  Use the checklist while making these observations.

DEBRIEFING

  The Audit Team Leader should first meet alone with the audit
team members to review their results and determine  what should
be addressed  at the debriefing.  The review should address the
following points, allowing team members to summarize their find-
ings:
•Sampling activities and documentation
•Sampling bank activities and documentation
•QA problems
•Follow-up on previous recommendations
•Summary
  Debriefing should be  held  between the  audit team and  project
personnel deemed appropriate by the Project Officer.
  In most cases, the Team Leader should conduct the debriefing
and review the team's iniiial  findings. The Leader may choose to
let team members comment on  their own findings.  It should be
made clear that the results of the audit are still tentative at this stage
and that the final audit results will be reported in writing.
  After each topic is discussed,  allow project personnel to make
comments. The  Team Leader should request any further docu-
mentation he may need  for the final report; resumes of new peo-
ple, copies of additional protocols, etc.
  Each  team member should write a report  on his findings. This
should include a copy of the completed  checklists.  The  reports
are  then assembled into  a  consensus  document by  the Team
Leader  and  reviewed by the  Audit  Program Manager. The final
revised  report, signed by the  Team  Leader and approved  for dis-
tribution by the Program Manager, is released to the office request-
ing the audit and the Project Officer.1
  The  report should include  and clearly identify points which re-
quire corrective action. These should be in the form of recommen-
dations.
ACKNOWLEDGEMENTS

  This Paper has been reviewed in accordance with the U.S. En-
vironmental Protection Agency's peer and administrative review
policies and approved for presentation and publication.
REFERENCES

1. USLFA, "Interim Guidelines and Specifications for Preparing Quality
   Assurance Project Plans,"  U.S. Environmental  Protection Agency,
   QAMS-005/80, 1980.
2. Barth, D.S.,  and Mason, B.J., "Soil Sampling Quality Assurance
   User's Guide,"  U.S.  Environmental  Protection  Agency, Las  Vegas,
   NV.USEPA 600/4-84-043, 1984.
3. Owens, T.W., "Standard Operating Procedures for conducting Sam-
   pling Team and Sample Bank Audits" Life Systems,  Inc. In Review
   U.S. Environmental Protection Agency, Las Vegas, NV.
4. F.C.  Hart, Associates, Inc.,  "RCRA Inspection  Manual," U.S. En-
   vironmental Protection  Agency Office of Solid  Waste and  Emer-
   gency Response,  1981.
5. USEPA, "Safety Manual for Hazardous Waste Site Investigations,"
   U.S. Environmental Protection Agency-National  Enforcement Inves-
   tigations Center, 1979.
6. USEPA, "Health and Safety Requirements lor  Employees Engaged In
   Field Activities," USEPA Order 1440.2,  U.S.  Environmental Protec-
   tion Agency, Washington, D.C., 1981.
 96
         AIR MONITORING

-------
   MODELING MOBILIZATION  AND  FATE  OF LEACHATES
     BELOW  UNCONTROLLED  HAZARDOUS WASTE  SITES

                                    MARCOS BONAZOUNTAS,  Ph.D.
                                            Arthur D. Little, Inc.
                                          Cambridge, Massachusetts
                                              JANET WAGNER
                                             DIS/ADLPIPE, Inc.
                                          Cambridge, Massachusetts
INTRODUCTION
  There are four major exposure pathways for contaminants from
uncontrolled  hazardous  waste  disposal  sites:1  ground-
water/leachate, surface water, contaminated soils  and residual
waste and air (Figs. 1 and 2). The environmental setting for an un-
controlled disposal site located above the water table is shown in
Figure 1. The potential pathways to human and ecological recep-
tors of Figure 1 are depicted in Figure 2. A variation of the above
would be a site where the waste was buried below the water table; in
this case, the leachate flume and groundwater are coincident. The
exposure pathways are essentially the same in both cases.
  Remedial actions are designed to reduce exposure to humans and
the environment to acceptable levels either by containing pollutants
originating from the waste site in place  or by removing the hazar-
dous substances from the immediate environment.
  Modeling can play an integral role in waste cleanup and other en-
vironmental protection studies. A model is a decision tool which, if
applied properly, can greatly assist decision-makers in effectively
dealing with complex issues at uncontrolled waste sites. Today, five
basic model categories are used: (1) emission models, to quantify
release (quantity) or pollutant emissions in the environment (e.g.,
air emissions or leaching  from a  waste site; (2) fate models,  to
estimate concentrations of pollutants in the environmental media
(e.g., fate of pollutants in the soil and groundwater); (3) exposure
models, intended to convert environmental concentrations to ab-
   Vapors
                                                 Municipal
                                               Water Supply
                            " I
                      Leachale Plume
                        Figure 1
 Environmental Pathways from a Generalized Hazardous Waste Site
sorbed  doses by humans (e.g., in inhalation); (4) risk models,
known  also as dose-response  models, for  the extrapolation  of
animal carcinogenicity data to humans and the estimation of prob-
able human risks to cancer;  and (5) cost/effectiveness models  or
analyses (e.g., mathematical  optimization models) to estimate ef-
fectiveness (e.g., reduction of human risk) when imposing alter-
native actions or strategies (e.g., remedial actions) at waste sites.2
  In this paper, the authors  present information on leachate fate
modeling in the soil and groundwater regimes below uncontrolled
hazardous waste sites.

SOURCES, EMISSIONS, ENVIRONMENTAL
PHASES
  Soil and groundwater contamination are commonly encountered
problems at uncontrolled hazardous waste sites; they result from
the migration of leachates originating from a wide variety of waste
management facilities including storage and  treatment, landfills,
surface impoundments, mines, waste piles  and land  treatment.
Primary leachate and pollutant  sources and waste modes are given
in Table 1 and the sources and the associated pollutants are listed3
in Table 3.  The composition of the leachates produced depends
principally on the type of wastes present and the decomposition in
the waste site (aerobic or anaerobic).
  Modeling mobilization of leachates from waste sites is a complex
task; most  often,  the problem cannot be approached from a
simplified perspective, for example, by employing a one- , two-  or
three-dimensional model that accounts for convection,  dispersion,
adsorption,  retardation of decay of species in the soil  or ground-
water regime.  Pollutant species of the leachates partition in the
various phases of  the soil matrix (Fig. 3),  whereas the species
phases, the environmental dynamics and the species chemistry are
interactive at all times. This interaction governs the leachate migra-
tion and mobilization  in  the soil  compartment, especially since
secondary compounds  are produced in the various phases of the
soil matrix.
  For example, landfills are principally disposal sites for municipal
refuse and some industrial wastes. Municipal refuse is generally
composed of 40 to 50% (by weight) organic matter, with the  re-
maining mass consisting of moisture and inorganic  matter such as
glass, cans, plastic, pottery, etc. Under aerobic decomposition, car-
bonic  acid  that is  formed  reacts  with any metals present and
calcareous materials in the rocks and soil, thus increasing the hard-
ness and  metal content of  the leachate. Decomposition of the
organic matter also produces gases, including CO2, CH4, H2S, H2,
NH3 and N2, of which CO2 and CH4 are the most  significant soil
contaminants.3
                                                                                      LEACHATE CONTROL
                                                      97

-------
                HUMAN
                ixpoumi
                PATHWAYS
              INVlAOMKUMlAL
       Ditctiplion o' EapOiUM P*th*Myi

       1  InruUtion ol vapoit
       2. lotuUliOn ol ptrttcuUlu cont.
       3  S***ttowing ol 1*1911 fttMiclct

       4  litytition of lOil on v\w (jWMtcuUily childiwtl

       5  Intuit|ion ol ««|KM| tiom M*p*9« m t
       6  Irttfomon of toil oil utt (JMIiicuUMy

       7. Inyviliori o' flfouruhwltr
       8. hellion of planu (homt e*rd*ni)
                                      0. Dnmal COHIKI wiih ioil ovwt*
                                     10 0«tm»l cuf lici «""' "»•' u" »•'•
                                     11 Diimil conucl liurti huuwhukl UM ot frounft«v«l«f Ruling, tu I
                                     I? Dmmtl cwilMI */»*nng" "• b«wnwini
                                     11 lnoitl.unut write* w«i»i
                                     H D«im«l conlMl wild turtle* «•'•' Imtnvtimf. vu I
                                     Ib Inyvtirod ut «()u«ti« txou


                                                     Figure 2
                               Exposure Pathways from a Hazardous Waste Disposal Site
     Thus, modeling of mobilization and leachate below uncontrolled
  hazardous waste sites requires a thorough knowledge of the envir-
  onmental factors and chemistry of the site.
 ENVIRONMENTAL FACTORS AND CHEMISTRY

   Soil zone modeling is a complex problem. A major characteristic
 of a soil subcompartment—as contrasted to water or air subcom-
 partments—is  that  the temporal,  physical  and  the  chemical
 behaviors  of  this  subcompartment are  governed by "out-
 compartmental" forces such as precipitation, air temperature and
 solar radiation. This governance by external factors is also one of
 the main reasons why the mathematical modeling of  leachate
                           Figure 3
               Schematic of Phases in Soil Matrix
                                                           migration in soil can be much more complex than effluent or air
                                                           modeling.
                                                             The chemical, physical and biological properties of a leachate in
                                                           conjunction  with the environmental characteristics of an area,
                                                           result in physical, chemical and biological processes associated with
                                                           the  transport  and transformation of  the  leachate  in soil  and
                                                           ground water. These processes are described in the following  sec-
                                                           tions, along with some of the mathematical models described in the
                                                           literature.

                                                           Physical Processes
                                                             The physical behavior of a chemical determines how the chemical
                                                           partitions among the various  environmental media; this partition-
                                                           ing has a significant effect on  the environment fate of a substance.
                                                           For example,  the release into  soil of  two different acids (with
                                                           similar chemical behavior) may result in one chemical volatilizing
                                                           into the air  and the other chemical adsorbing onto the organic
                                                           material in  the soil.  The  physical behavior of a  substance,
                                                           therefore, can have a significant effect on the environmental fate of
                                                           that substance.
                                                             The processes and corresponding  physical parameters that are
                                                           important in determining the behavior and fate of small amounts of
                                                           chemicals differ from  forces governing chemical migration during
                                                           large-scale releases (e.g., spills).
                                                             The processes of advection,  diffusion, sorption and  volatiliza-
                                                           tion are most important to both trace-level analyses and large-scale
                                                           release analyses. Bulk  properties (e.g., viscosity and solubility) are
                                                           usually only important in simulations involving large amounts of
                                                           contaminants.

                                                           Sorption/'Ion-Cation Exchange
                                                             Adsorption  is  the adhesion  of  leachate   pollutant ions or
                                                           molecules to the surface  of soil solids,  causing an increase in the
                                                           pollutant concentration on the soil surface over the concentration
                                                           present in  the  soil moisture.  Adsorption occurs  as  a result of a
                                                           variety of processes with  a variety of mechanisms, and  some pro-
                                                           cesses  may cause an increase of pollutant concentration  within the
                                                           soil solids—not merely on the soil surface. Adsorption and desorp-
                                                           tion can drastically retard leachate migration in soils;  therefore,
                                                           knowledge of this process is of importance when one is dealing with
                                                           contaminant transport in soil  and groundwater. The type of pollu-
                                                           tant will determine to what  kinds of material the pollutant  will
                                                           sorb. For organic compounds, it appears that partitioning between
                                                           water and the organic  carbon  content of soil is the most  important
                                                           sorption mechanism.
98
LEACHATE CONTROL

-------
                                                                      Table 1
                                                                Sources and Wastes
Solid Waste Wastewater- Injection
Wastewater Disposal Spray Land or Disposal Septic Tanks
Pollutant Source Impoundments Sites Irrigation Application Wells and Cesspools
Industrial
Wastnwater X XX
Slutlgu ' X
Solid Wastu X
Municipal
Wastewater X
Sludge X X
- Solid Waste- X
Household
- Wastewater X
Agricultural Feedlot
Mining XX X
Petroleum Exploration X
Cooling Water X
Buried Tanks,
Pipelines
Agricultural
Activities
Infiltration/
Surface Leaching from
Pits Runoff Storage Sites



X



X


X
X X
X


X

X X
   Sorption and desorption are usually modeled as one fully reversi-
ble process, although Jiystersis is sometimes observed. Four types
of equations are commonly used to describe sorption/desorption
processes: Langmuir,  Freundlich,  overall and ion or cation ex-
change.  The Langmuir isotherm model was  developed for single
layer adsorption and is based on the assumption that maximum ad-
sorption corresponds to a saturated monolayer of solute molecules
on the adsorbent surface, that the energy of adsorption is constant
and that there is no transmigration of adsorbate  on the  surface
phase. These  models are thoroughly described  in  the literature;4
therefore, no additional informatioan is provided here,
   Ion exchange (an important sorption mechanism for inorganics)
is viewed as an exchange with some other ion that initially occupies
the adsorption site on the solid. For example, for metals (M + + ) in
clay the  exchanged ion is often calcium.
        M++ +  [clay]  •  Ca- Ca++ + [clay] •  M            (1)
Cation exchange can be quite sensitive to other ions present in the
environment. The calculation of pollutant mass immobilized by ca-
tion exchange is given by:
       S  = EC • MWT/VAL                                   (2)
where: S  = maximum mass associated with solid (mass pollutant/
mass of soil); EC = cation exchange capacity (mass equivalents/
mass of dry soil); MWT  = molecular (or atomic) weight of pollu-
tant (mass/mole);  VAL =  valence of ion (-). For additional
details, see the article by Bonazountas and Wagner.5
Diffusion/Volatilization
   Diffusion in solution and  volatilization from the soil-air to the
atmosphere are processes that  affect leachate migration of mainly
volatile compounds. Many volatilization models are  available in the
literature, but some of these models can be applied only to specific
environmental situations.4

Chemical Processes
   The important chemical processes to be considered when model-
ing mobilization and fate of leachate are: ionization, hydrolsys, ox-
idation/reduction  and complexation.
Ionization
   Ionization  is  the  process of separation or  dissociation  of  a
molecule into ions—particles of opposite electrical charge. The ex-
tent of ionization has a significant effect on the chemical behavior
                            Table 2
Primary Sources of Soil Contamination and Associated Pollutants
      Source

 Industrial Sources

 Chemical manufacturers

 Petroleum refineries

 Metal smelters and refineries

 Electroplaters

 Paint, battery manufacturers

 Pharmaceutical manufacturers
 paper and related industries

 Land Disposal Sites

 Landfills that received sewage sludge,
 garbage, street refuse, construction
 and demolition wastes


 Uncontrolled dumping of industrial
 wastes, hazardous wastes

 Mining Wastes


 Agricultural Activities


 Agricultural feedlots

 Treatment of crops and/or soil with
 pesticides and fertilizers; runoff or
 direct vertical leaching to septic
 tanks and cesspools


 Leaks and Spills

 Sources include oil and gas wells,
 buried pipelines and storage tanks;
 transport vehicles

 Atmospheric deposition
 Highway Maintenance Activities

 Storage areas and direct application

 Radioactive Waste Disposal

 Eleven major shallow burial sites exist
 in U.S.; 3 known to be leaking

 Land Disposal of Sewage and Wartewater

 Spray irrigation of primary, secondary
 effluents


 Land application of sewage sludge

 Leakage from sewage oxidation ponds
     Type of Pollutant!


Organic solvents

Chlorinated hydrocarbons

Heavy metals

Cyanide, other toxics

Conventional pollutants

Acids, alkalines, other corrosives
many are highly mobile in soil.


BOD, inorganic salts, heavy metals
pathogens, refractory organic compounds,
plastics; nitrate; metals including iron,
copper, manganese suspended solids
Acidity, dissolved solids, metals, radio-
active materials, color, turbidity

BOD, nutrients, fecal coliforms, chloride,
some heavy metals
Herbicides, insecticides, fungicides,
nitrates, phosphates, potassium, BOD,
nutrients, heavy metals, inorganic salts,
pathogens, surfactants; organic solvents
used in cleaning
Petroleum and derivative compounds;
any transported cnemicals


Particulates; heavy metals, volatile
organic compounds; pesticides; radio-
active particles


Primarily salts


Primarily 13SCS, 90Sr, and *°Co
BOD, heavy metals, inorganic salts,
pathogens, nitrates, phosphates,
recalcitrant organics
                                                                                                             LEACHATE CONTROL
                                                                                                                                                99

-------
of a chemical in a leachate. An acid or base that is extensively ion-
ized may have markedly different solubility, sorption, toxicity and
biological characteristics  than  the  corresponding  neutral  com-
pound. Inorganic and organic acids, bases and salts may be ionized
under environmental conditions. A weak acid (HA) will ionize to
some extent in water according to the reaction:
                           +A-                           (3)
The acid dissociation constant Ka is defined as the equilibrium con-
stant for this reaction:

       Ka = [H30 +][A -]/[HA][H20]                     (4)
Note that a compound is 50% dissociated when the pH of the water
equals the pKa (pKa  = -  log Ka).

Hydrolysis
   Hydrolysis is one  of a  family of reactions which transforms a
leachate pollutant. Under environmental conditions existing at a
waste site, organic compounds are the main chemicals hydrolyzed.
Hydrolysis  is a  chemical transformation  process  in  which an
organic (RX) reacts with water, forming a new molecule. This pro-
cess normally involves the  formation of a new carbon-oxygen bond
and the clearing of the carbon-X bond  in the original molecule:

       RXHzOR-OH +  X-  + H+                        (5)

Hydrolysis reactions are usually modeled as first-order processes
using rate constants (KH) in units of (time.)'

       -d[RX]/dt = KH[RX]                               (6)

The  rate of hydrolysis  of various organic chemicals,  under  en-
vironmental conditions, can differ by 14 orders of magnitude with
associated half-lives as low as a few seconds to as high as 106 years.
If laboratory rate constant data are used in soil models and not cor-
rected for environmental  conditions, as it  often the case, then
model results should be evaluated with  skepticism.

Oxidation/Reduction
   For some organic  compounds of leachates (such as phenols,
aromatic amines, electron-rich olefins  and dienes,  alkyl sulfides
and eneamines), chemical oxidation  is an important degradation
process. Most of these reactions depend on reactions  with free-
radicals  already in solution and are  usually modeled by pseudo-
firs t-order kinetics:

      - d[X]/dt = K'0 [R02 • ]  [X] =  KO, [X]                 (7)
where: X is the pollutant, K'0 is the second order  oxidation rate
constant, RO2= is a free radical and Kox is the pseudo-first-order
oxidation rate constant.
Complexation
   Complexation, or  chelation, is the process by which metal ions
and organic or other non-metallic molecules (called ligands)  can
combine to  form stable metal-ligand complexes. The complex that
is found will generally prevent the metal from undergoing other
reactions or interactions that the free metal cation would undergo.
Complexation may be important in some situations; however, the
current level of understanding  of the process is not very advanced,
and the available information has not been shown to be particularly
useful to quantitative modeling.1

Biological Processes

   Bioaccumulation is the  process by which terrestrial  organisms
(such as plants and soil invertebrates) accumulate and concentrate
pollutants from the soil. Bioaccumulation is not examined in  soil
modeling,  aside  from  some  nutrient  cycle  (phosphorus and
nitrogen) and carboncycle bioaccumulation attempts.
   Biodegradation refers to the  process  of transformation of a
chemical by biological agents, usually  by microorganisms. It in-
cludes a number of different  processes  such as: mineralization,
detoxication, cometabolism, activation and change in spectrum. In
toxic chemical modeling,  biodegradation is usually  treated as a
first-order degradation process.'
       dc/dt = - KDE * cn                                  ^8^
where:  c  = dissolved concentration of pollutant soil moisture
(/ig/mL);  KDE  = rate of degradation (day-1); and n = order of
the reaction (n  = 1; i.e., a first order reaction).
MATHEMATICAL MODELING

  Leachate  and  pollutant fate  mathematical modeling  in soil
systems is an  area of current  intensive  work  because  of the
numerous problems originating at hazardous waste sites. The varie-
ty of models has dramatically  increased during the last decade.
Although  the numbers of models appears to be large, only a very
few different modeling concepts exist and very  few physical or
chemical processes are modeled.
  In general, soil/groundwater modeling concepts deal mainly with
point source pollution and can  be categorized as: (1) unsaturated
soil zone (or soil), (2) saturated soil zone (groundwater), and (3)
geochemical. The first two categories follow comparable patterns
of mathematics and approach; the third enters into chemistry and
speciation modeling as presented by Bonazountas.'
  One effective way to account  for all previously reported pro-
cesses in a modeling  study is to formulate a "compartmcntal"
model that assumes pollutant equilibrium at all phases and between
all compartments of the soil matrices shown in Figure 3. This for-
mulation is achieved by applying: (1) the law of pollutant mass con-
centration for a representative species of the leachate  over time, in
each phase,  in each compartment and for all processes (physical,
chemical, biological) and (2) to seek a balance of pollutant masses
of the various compartments at all times. The chemical dynamics of
the model have to be supplemented with the hydrogeologic and,
perhaps, the watershed dynamics of the overall compartment.
  At this stage of scientific research, the most developed soil  com-
partment model appears to be  SESOIL: Seasonal Soil Compart-
ment  Model.' SESOIL is a user-friendly, mathematical soil  com-
partment model designed for long-term environmental, hydrologjc,
sediment and pollutant fate simulations. It  can  describe:  water
transport (quality and quantity),  sediment  transport  (quality and
quantity),  pollutant transport and  transformation,  soil quality,
pollutant migration to groundwater and other processes. Simula-
tions are performed for  a user-specified soil column extending be-
tween the ground surface and the lower part of the  unsaturated soil
zone of a region (Fig. 4).
  The simulations are based upon a three-cycle rationale, each cy-
cle being associated with a number  of processes. The three cycles
are: (1) the hydrologic cycle which  takes account of rainfall, soil
moisture,   infiltration,   exfiltration,  surface  runoff,  evapor-
transpiration and groundwater runoff; (2) the sediment cycle which
takes account of sediment washload (from storms) and sediment
resuspension (due to wind); and (3) the pollutant fate cycle which
takes account  of advection, diffusion,  volatilization, adsorption
and desorption, chemical degradation or decay, biological transfor-
mation, hydrolysis, photolysis  (not  operational), oxidation (not
                           Table 3

Compound
Chromium
Copper
Sodium
Calculated Amounl
of Compound
Applied
(April 197V)
2.92 M»/cm2
3.24 jig/cm2
8.64 v 103 Bg/cm2
Ci
Measure
0-1 5cm 1
0.80
0.20
0.89
                                    Concentration Measured (ppra)
                                       Unlj I«T»)

                                                  Prrfklrd
                                     15-JO cm    O-lSctn      I WO cm
                                       1.0
                                       0.0
                                      114.0
 O.IS
 0.16
85.6
 0.005
 00003

142.0
100
          LEACHATE CONTROL

-------
    Cvaporat't
                           Figure 4
       Schematic Presentation of the Soil Compartment (Cell)
operational), complexation of metals by organic ligands, cation ex-
change, fixation (not operational), nutrient cycles (not operational)
and other processes. Model development has been sponsored by the
USEPA. SESOIL model application and validation studies by its
developers have been undertaken for land treatment practices,5 for
human exposure assessment  studies related  to groundwater con-
tamination and for the fate of volatile solvents in soil systems.6 The
model has been tested and validated by many researchers.

CASE STUDIES
  Two studies were performed on leachate mobilization and migra-
tion at uncontrolled waste sites: (1) leachate migration to ground-
water from land treatment practices' and (2) leachate migration to
the air of solvents leaking from barrels buried in the soil zone.'

Leachate to Groundwater—Land Treatment Practices

  The land treatment site considered is the property of a  plastics
manufacturing plant. Manufacturing process wastes are treated in a
secondary wastewater treatment system at the facility. Sludge from
the wastewater treatment system is centrifuged to yield a sludge
whose content is 5-10"% solids; the resulting sludge is disposed of by
land treatment.
  In July,  1979, 5400 kg/ha of sludge were incorporated into the
soil of a clean (i.e., not previously land cultivated) area of the site.
The sludge was injected 12.7 to 20.3 cm below the soil surface and
was subsequently mixed with the soil by ordinary farming methods.
  The soil in the land treatment area is silt-loam, with a spatial in-
trinsic permeability of 7.05 x  10~9 cm2 and a surface slope of 3%.
Depth of groundwater is reported  to be 30 to 70 m. The  40-year
(1940-80) average annual rainfall is about 85 cm. The seven year
rainfall (1973-80) is about 84 cm and last year's (July 79-August 80)
rainfail was 79 cm. The average time of rain varies between 0.18
and  0.20  day for the  above periods.  The  area receives  84-110
rainstorms per year. The rainy season is 365 days per year. The an-
nual average temperature is 14 °C. Almost no surface runoff occurs
at the site, due to both the climatic and the soil conditions.
  Waste application occurred in the spring of 1979. In July, 1979,
and a year later in August, 1980, soil core samples were collected
from 2 depths, 0 to 15 cm below grade and 15 to 30 cm below grade
at both the waste application area and at a control area. The con-
trol area soil was nearly identical to the soil of the waste application
area, except that no waste had been applied.
   Soil core samples were air-dried prior to analysis. Analyses were
performed on nitric-percloric.acid digests of representative aliquots
of the respective soil samples, so that reported results represent
total metal concentrations and do  not differentiate between ad-
sorbed and dissolved analyte. Laboratory results are expressed as
micrograms of analyte per gram of air-dried soil. Laboratory
analysis of samples collected in August, 1980, is not yet complete.
   Chemical data have been obtained from the literature and from
site-specific investigations. No calibration has been attempted for
the  hydrologic cycle routine or for the  soil parameters.  The
chemical parameters (adsorption coefficients) have been adjusted
(by up to 20%) to calibrate the results.
   The SESOIL was used to predict average concentrations of in-
organic pollutants  (Table  3).  Predicted  concentrations  agree
reasonably well with those values measured chemically, considering
the uncertainty of all parameters affecting pollutant transport in
soil  compartments.  A sensitivity analysis is being performed  to
study impacts of changes upon the  soil compartment quality: (1)
sludge application rates, (2) climatologic  and  soil parameters and
(3) chemistry parameters. Results of the analysis are presented
elsewhere.5

Leachate to the Atmosphere—Buried Solvents
   The purpose of this research is twofold:  (1) to understand—using
a  mathematical simulation—the long-term potential fate of the
leachate of six solvents leaking from buried barrels disposed in soil
systems, (2)  to test the performance of  the  SESOIL model for
highly volatile compounds. For this investigation, six halogenated
organic solvents have been examined.'
•Perchloroethylene (Tetrachloroethene)
•Methylchloroform (1,1,1-Trichloroethane)
•Methylene Chloride (Dichloromethane)
•Carbon Tetrachloride (Tetrachloromethane)
•Freon  113
•Trichloroethylene (1,1,2-Trichloroethane)
   It was not the authors' intention to conduct a site specific study ;
therefore, a  number  of hypothetical scenarios covering a wide
range of U.S. climates, soils and solvents were considered. The
methodology developed for the overall fate assessment is of general
use and can be employed for similar studies or classes of pollutants.
  The major  conclusions of this analysis are shown in Table 4 and
Figure 5. The fate of buried solvents was determined by simulation
runs performed using SESOIL to determine the approximate pollu-
tant  mass of organics volatilizing, migrating to the groundwater or
being entrained in the soil column after a period of ten years. This
analysis yielded the following conclusions:
•Of all the chemicals studied, Freon 113 is most easily volatilized,
 whereas  methylene chloride  is  least easily  transported to  the
 atmosphere.
•Methylene chloride contributes the most mass to  groundwater;
 Freon  113 contributes the least pollutant mass to groundwater.
•The other solvents have fates  intermediate  between Freon 113
 and methylene chloride and are fairly  similar to one another.
 Under moderate conditions, 99-64% of their mass  volatilized
 and 0.01-3% of their mass reached the groundwater. The remain-
 ing mass was captured in the soil column.
•Leaching to  groundwater  increases for chemicals  with low
 Henry's  law constants, low diffusion coefficients and low ab-
 sorption coefficients. Leaching is generally favored by high rain-
 fall and permeable soils.
•Volatilization  is favored for chemicals with high  Henry's law
 constants and high diffusion rates.  It is generally enhanced by dry
 conditions and porous soil. Decreasing soil column depth gen-
 erally  results in increasing volatilization rates up to a certain
 depth. In this model version, volitilization appears to be the pre-
                                                                                             LEACHATE CONTROL
                                                          101

-------
                  VOLATILL-. COMPOUNDS MODHLING
        ?raa!c*.ee F«t» of  Solvents

~
„..
I "i
s
! '7
i ,.-
E so-
ft
1 "T
1 4"

...
10- •

— 90 Naur*
--11 man
-



r-|




1

i iJ . . .










r-

!













!
^ .


















-






!
1












1-1
i
J
f



r
i





i
i
«
laaaj 1



1








.
1*1 *
n'
u. . .



-







s
i
,












P
i












i



—








i
i













u
3 t
-,
       Ranoi gf Solvent Fataa
        leoln: 50 M«t«r»
   I  W
   !

   \

   \
                                     \~
                            Figure 5
     Predicted Ranges of Solvent Fates—Summary of Simulations

dominant removal process from the soil for all  the  chemicals
studied.
•The actual quantities of  mass removed by each  pathway  are
 strongly affected by the climate and soil type. Summaries of the
 pathways   for  all the  six chemicals  for  a  10-year  simulation
 period,  a moderate climate, a silty loam soil and three depths of
 the soil column are given in Table 4.  The range of all pollutant
 fates for all solvents considered and one typical climate, one  soil
 type and two soil column depths is shown in Figure 5b. The fates
 of three solvents for all scenarios considered in this study (includ-
 ing sensitivity  analyses  of important variables), are  shown in
 Figure 5b.

CONCLUSIONS

  Mathematical modeling is  an essential  and powerful tool  for
assessing the mobilization and fate of leachate below uncontrolled
hazardous waste sites. Models exist in the literature; however, they
have to be appropriately selected and applied. Models have to ac-
count for the physical, chemical and biological processes of a site,
although exact  knowledge of the physics  of the  soil system—al-
though  essential—is impossible  prior  to employing any model.
Model output validation is essential  to any  soil modeling  effort,
although this term has a broad meaning in the literature.
                                                             For the purpose of this paper, the authors have defined valida-
                                                           tion as "the process which analyzes the validity of final model out-
                                                           put," namely the validity of the predicted pollutant concentrations
                                                           or mass in the soil column (or in groundwater) to groundwater and
                                                           to the air as compared to available knowledge of measured pollu-
                                                           tant concentrations  from  monitoring data (field sampling).  A
                                                           disagreement  of  course  in  absolute levels of  concentration
                                                           (predicted versus measured) does not necessarily indicate that either
                                                           method of obtaining data (modeling or field sampling) is incorrect
                                                           or that  either data set needs  revision.  Field sampling approaches
                                                           and modeling approaches rely on two different perspectives of the
                                                           same situation.
                                                                                      Table 4
                                                                             Quantitative File of Solvent**
                                                                           Che»lcal

                                                                      Depth to CroundMeter; 50 Metera

                                                                      Tetrachloroethene
                                                                      I, ,1-TrIchloroethane
                                                                                                  1        1 Meaalnlng     1 Leacned
                                                                                              Voletlllied  In Soil Colu«n to Crowndweter
                                                                        hylene Chloride
                                                                        bon Tetrechlor Ida
                                                                        on 113
                                                                        chloro*thane
                                                                                       7«.9

                                                                                       37!«
                                                                                       62.•
                                                                                       98.5
                                                            Depth to Groundwater!  20 Hetera

                                                            Telrechloroethene              B8.0
                                                            1,1,1-TrIchloroethane           9*.2
                                                            Hethylene Chloride             57.0
                                                            Carbon Tatrachlorlde           9'.3
                                                            Freon 113                    99.6
                                                            Trlchloroethene               82.4
                                                            Depth to Croun
                                                            Tetrachloroethene
                                                            1,1,1-TrIchloroethane
                                                            Hethylene Chloride
                                                            Carbon Tetrachlorlde
                                                            Freon 113
                                                            Trlchloroethene
99.3
;•  2
99.3
M.9
96.0
             2«.7
             17.3
             60.2
             17.2
              1.5
             3«.7
                                                                                                             10.6
                                                                                                              • .9
                                                                                                             33.8

                                                                                                              o!s
 1. I
 0.1
10.3
 0.1
 0.01
 1.6
               0.3
               2.5
               0.3
              00.1
               0.9
                            1.3
                            0.9
                            9.2
                            o.e
                            0.01
                            3.0
1.1
0.6
6.5
0.6
0.01
2.3
                                                                      1. Percentage of aiasa after 10 yea^s. Boderata cllaatc. sllty lo
                                                                        Totals nay not add to 10O1.
                                                           • Percentage of mass iftet 10 jrcau, moderaic dinule. n»ly lo*m toU. Toub may not add to 100*.


                                                           ACKNOWLEDGEMENTS

                                                             To Mr. Michael Callahan, Dr. William Wood, and Dr. Annette
                                                           Nold, and M.W. Slimak, USEPA, Office of Toxic Substances, the
                                                           authors  give  their  gratitude  for  their  support and  technical
                                                           guidance during conductance of this research.

                                                           REFERENCES

                                                           1. Ehrenfeld,  J.  and  Bass, J.,  Handbook for  Evaluating Remedial
                                                             Action  Technology  Plans. USEPA Report No.  EPA-600/2-83-076,
                                                             Aug.. 1983, USEPA Cincinnati, OH.
                                                           2. Bonazountas,  M.  and Fiksel,  J., ENVIRO: Environmental Mathe-
                                                             matical Pollutant Fate Modeling Handbook/Catalogue. USEPA Con-
                                                             tract No. 68-01-5146. Arthur D. Little, Inc., Cambridge, MA, 1982.
                                                           3. Bonazountas, M., "Soil  and Groundwater Fate Modeling," in Fate
                                                             of Chemicals in the  Environment. R. Swann and A. Eschenroeder,
                                                             eds.,  ACS  Symposium Series  No. 25, American Chemical  Society,
                                                             Washington, DC,  1983.
                                                           4. Bonazountas,  M.  and Wagner, J.,  SESOIL, A Seasonal Soil Com-
                                                             partment Model. USEPA/OTS Contract No. 68-01-«271, Arthur D.
                                                             Little, Inc., Cambridge, MA, 1984.
                                                           5. Bonazountas, M., Wagner, J.  and Goodwin, B., Evaluation of Sea-
                                                             sonal Soil/Groundwater Pollutant Pathways via SESOIL.  Final Draft
                                                             Report, USEPA/MDSD Contract No. 68-01-5949(9), Arthur D. Little,
                                                             Inc., Cambridge, MA, 1984.
                                                           6. Wagner, J. and Bonazountas, M., Potential Fate of Halogenated Sol-
                                                             vents via SESOIL, First Draft Document, USEPA/OTS Contract No.
                                                             68-01-6271, Arthur D. Little, Inc., Cambridge, MA,  1983.
102
LEACHATE CONTROL

-------
IMPLEMENTATION  OF A COOPERATIVE  AGREEMENT TO
            INVESTIGATE AND  REMEDY  SURFACE AND
  GROUNDWATER CONTAMINATION  AT THE BOULDER/
 MARSHALL LANDFILL,  BOULDER COUNTY,  COLORADO
                                        PAUL V. ROSASCO
                                        Fox Consultants, Inc.
                                          Denver, Colorado
                                         JOHN CURRY, JR.
                                      Browning-Ferris Industries
                                       Eden Prairie, Minnesota
INTRODUCTION

  The inactive Marshall Landfill located  in southeast Boulder
County, Colorado, has been designated as the 81st site on the Na-
tional Priorities List and is the highest ranked  Superfund Site in
Colorado. From approximately 1955 to  1974, both uncontrolled
and engineered disposal of solid waste, liquid sewage sludge and
septic pumpings and light commercial and  industrial wastes have
occurred at this facility. The adjacent Boulder  Landfill, an engi-
neered sanitary landfill accepting only solid waste, was opened in
1975.
  These facilities  (collectively referred to as the landfill) are lo-
cated on the flank of Lake Mesa, a large pediment capped mesa.
Natural seepage along the mesa flank has resulted in generation of
approximately 30,000 gal/day of heavy metal and organic leachate
which threatened  to impact Community Ditch, a large irrigation
and municipal water supply canal which traverses the inactive land-
fill. The landfill is also situated over a presumed recharge area for
a major regional aquifer of the Denver  Basin. Heavy metal and
organic contamination has been detected in on-site and peripheral
monitor wells within the alluvial aquifer.
  In 1983, a Cooperative Agreement was signed between involved
State and local regulatory agencies and impacted parties to investi-
gate and remedy contamination at the landfill.  In this paper, the
authors discuss site conditions, the background and requirements
of the Cooperative Agreement and the results obtained to date.
 BACKGROUND
  The Boulder/Marshall Landfill is located in southeast Boulder
 County, Colorado, approximately midway between the towns of
 Marshall and Superior, approximately seven miles southeast of the
 community of Boulder, Colorado and 20 miles northwest of Den-
 ver, Colorado.

 Disposal History

  The landfill consists of four distinct tracts (Fig. 1):
 •An abandoned landfill operated as an open dump from approx-
 imately 1955 to 1970.
 •The inactive Marshall Landfill operated by  Urban Waste Re-
 sources from 1970 to 1974.
 •Associated septic pumpage disposal ponds operated by  Urban
 Waste Resources from approximately 1970 to 1975.
 •The active Boulder Landfill and associated gravel mining opera-
 tion, first opened in 1975 by Urban Waste Resources and subse-
 quently sold to Landfill Inc. in 1976.
  Landfill Inc. subsequently permitted an above grade expansion
which is expected to keep the active Boulder Landfill in operation
until 1990.
  The Boulder/Marshall Landfill is located along the  flank  of
Lake Mesa, a broad upland pediment surface. Refuse in the active
landfill is placed across the top of Lake Mesa.  Refuse within the
inactive landfill was placed along the flank of Lake Mesa, down
into the Cowdrey Drainage and up along the bedrock slope north
of the drainage.
  Although the landfill was  intended primarily for domestic,
municipal and light commercial solid  wastes only, several other ma-
terials  including untreated municipal sewage sludge from the City
of Boulder were also disposed of at the landfill. From 1972  to
1974, while the city's waste treatment plant was being upgraded,
sludge was co-disposed with the solid wastes by excavating slit
trenches within previously placed solid wastes and emptying the
nearly 80% liquid sewage sludge directly into the refuse. Septic
tank pumpings were also dumped in a series of unlined open pits
                        Figure 1
             Site Plan of the Boulder/Marshall Landfill
                                                                               LEACHATE CONTROL
                                                                                                         103

-------
along the east side of the landfill between 1970 and 1975. Co-dis-
posal of septic pumpings with the refuse may also  have occurred.
Finally, records at the County Health Department indicate that in-
dustrial chemicals, primarily  organic solvents, were co-disposed
with the refuse until 1975.


Surface Water Systems
  The Boulder/Marshall Landfill area encompasses two major sur-
face water systems, the Marshall Lake-Community Ditch  system
and the Cowdrey  Reservoir No. 2—Cowdrey Drainage System.
Marshall Lake was created by the construction of Marshall Dam in
1909 to provide irrigation supply for farmers in eastern Colorado
and peak municipal demand for the City of Louisville. Inflow into
the  Lake is  derived  primarily  from Community Ditch  which
collects surface water from South Boulder Creek near the town of
Eldorado Springs approximately 4 miles to the west of the landfill.
Outflow from  the lake occurs as flow along  Community Ditch,
seepage beneath the dam into Cowdrey Reservoir No. 2, and  seep-
age into the alluvial and bedrock aquifers.
  From Marshall Lake, Community Ditch traverses  the southern
portion of the  inactive landfill  and then straddles the northwest
flank of Lake Mesa to a point one mile northeast of the landfill
where it joins  the Louisville  Ditch. From there the water flows
either into Louisville Ditch to the Louisville treatment  plant or con-
tinues flowing in Community  Ditch. Currently, flow  in Commun-
ity  Ditch  is limited to the late spring  and summer  months during
peak demand for irrigation water. Maximum  flow rate has  been
recorded at 86 cfs.
  Cowdrey Reservoir  No. 2 is supplied primarily  by seepage be-
neath Marshall Dam with a lesser contribution from  runoff  from
the surrounding area and seepage from the landfill. Outflow  from
Cowdrey  Reservoir No. 2 occurs primarily as surface flow through
Cowdrey  Drainage. Due  to the low  gradient of this drainage,
approximately  1 to 5 ft/1000 ft, the drainage does not possess a
well-defined channel.  Rather, it consists  of a series of stagnant
ponds and marshes which eventually drain into  South Boulder
Creek.

Hydrogeology

  Five distinct water-bearing units have been  identified beneath
the landfill. These include an alluvial aquifer and  four distinct
sandstone units.
  The alluvial  aquifer  consists of: sands, gravels and clays mant-
ling the top and flank  of Lake Mesa; colluvial material along the
base of Lake Mesa; weathered  bedrock in  the low areas around
Cowdrey  Reservoir No. 2 and  Cowdrey Drainage; and alluvium
along Cowdrey Drainage. Inflow into the alluvial aquifer in the
vicinity of the landfill generally results from groundwater recharge
along the top of Lake Mesa and subsequent lateral flow to the
north-northwest, entering the landfill area along the southern and
eastern boundaries. In  addition,  a minor contribution to flow be-
neath the landfill occurs along the central portion of the western
boundary of the active  landfill due to recharge  from Marshall
Lake. In 1975,  a French Drain was constructed along the west side
of the active landfill to intercept this flow.
  Discharge from  the aquifer occurs  primarily as surface seepage
along the slope above and on the south side of Community  Ditch
where the water table intersects the ground surface. When this in-
vestigation began, this  seepage either  collected in the two lagoons
and evaporated or flowed into Community Ditch; flowed over the
inactive landfill, via the Lagoon No. 1 to discharge pipe across the
ditch and eventually into Cowdrey Reservoir or Cowdrey Drain-
age;  or seeped directly into  Community  Ditch.  The remaining
flow in the alluvial aquifer exits the area as seepage into or sub-
surface flow along Cowdrey Drainage.
  Immediately  beneath the alluvial aquifer is a shallow bedrock
aquifer consisting  of thin sandstone layers. This aquifer is  sepa-
rated from the  overlying alluvial aquifer by thin organic rich  clays
and weathered claystone. In  addition to  the  shallow bedrock
                                                       aquifer, at least three other saturated zones were known to occur
                                                       immediately (100 to 200 ft depth) beneath the landfill. These are
                                                       the "A" and "B"  sands  of the lower Laramie  Formation and
                                                       the Milliken Sandstone of the Fox Hills. Together, these three units
                                                       make up the Laramie-Fox Hills (L-F) aquifer, a major regional
                                                       aquifer of the Denver  Basin. Flow within the bedrock aquifers is
                                                       complicated by extensive high angle faulting. Prior to this investi-
                                                       gation,  no information was available on the potentiometric sur-
                                                       faces of these strata, on their hydraulic interconnection, or their
                                                       relationship  with the  overlying alluvial and shallow  bedrock
                                                       aquifers.  However,  it  was known that the landfill  was situated
                                                       over  the  recharge area for the regional L-F aquifer. It was the
                                                       potential threat to this aquifer, along with the potential impact to
                                                       the municipal water  supply flowing in Community Ditch, that led
                                                       the Colorado Department of  Health to nominate the inactive
                                                       Marshall Landfill as the highest priority Superfund site in the State.

                                                       Water Quality
                                                         Previous water quality monitoring had indicated  that existing
                                                       leachate, surface water and groundwater contamination were dom-
                                                       inated  primarily  by elevated concentrations  of totaJ  dissolved
                                                       solids, iron, manganese and phenolic compounds. Secondary com-
                                                       ponents of surface and groundwater  contamination at the landfill
                                                       included barium, lead,  cadmium, arsenic, mercury; cyanide; ben-
                                                       zene; phthalate esters; and  various volatile  organics  including
                                                       chlorinated hydrocarbons such as dichloroethylene, methyl  chlor-
                                                       ide, methylene chloride, trichloroethyleneand tetrachloroethylene.
                                                         Although these substances  had been  detected in Community
                                                       Ditch immediately downstream of the landfill, none were detected
                                                       at the point of diversion to the Louisville water system or  in the
                                                       raw water storage for the Louisville water treatment plant. In addi-
                                                       tion, due entirely to a lack of monitoring,  no impact to the reg-
                                                       ional L-F aquifer had been documented.
                                                       THE COOPERATIVE AGREEMENT
                                                         To ensure involvement in any investigation  or  cleanup  at the
                                                       landfill, a Cooperative  Agreement to study and remedy contam-
                                                       ination at the landfill was developed and entered into on June 6,
                                                       1983 bet ween:
                                                         1. The Commissioners of Boulder County, Colorado
                                                         2. The Colorado Department of Health (CDH)
                                                         3. Farmers Reservoir and Irrigation Company (FRICO)
                                                         4. City of Louisville,  Colorado
                                                         5. Landfill Inc. (LI), the current operator of the active Boulder
                                                           Landfill
                                                         Not one of these parties was a generator of wastes disposed at
                                                       the landfill; however, all of them perceived potential damages re-
                                                       sulting from the existence or threat of surface or groundwater con-
                                                       tamination.
                                                         The  Cooperative Agreement was structured after the  National
                                                       Contingency Plan (10 CFR 300) and set forth a three-phased pro-
                                                       gram of investigation and remedial action:

                                                       Phase I—Initial Remedial Measures (IRM)

                                                       •Design and install a 60 in. pressurized raw  water  pipeline along
                                                        Community Ditch through the landfill area
                                                       •Design and construct a seepage control system to prevent surface
                                                        and groundwater contamination due to leachate seepage
                                                       •Design and construct a monitoring program  to evaluate the effec-
                                                        tiveness of these two initial measures
                                                       Phase II—Remedial Investigation (Rl)
                                                       •Complete an investigation of possible groundwater contamination
                                                        in the landfill area
                                                       Phase III—Feasibility Study (FS)
                                                       •If any groundwater contamination presenting a hazard to human
                                                        health or the environment is detected, identify any cost-effective
                                                        remedial measure
 104
LEACHATE CONTROL

-------
•Design a monitoring system to identify any sources of surface
 and/or  groundwater contamination which  are  not effectively
 contained, isolated or neutralized by the initial or final remedial
 measures.
  Funding for these actions was to be obtained through a complex
formula. Installation of the pipeline was to be funded in part by the
City of Louisville and in part by Boulder County. Louisville was
to contribute monies equivalent to those necessary to install a 27
in. diameter pipeline that had previously been proposed as a means
of protecting only its portion of the ditch waters. Boulder County
could draw from an existing environmental "sinking fund" created
from a surcharge on disposal fees at the landfill and monies con-
tributed by the current operator. This would fund the difference in
the costs between  the 27 and the 60 in. pipelines. FRICO was
responsible for all  maintenance and repair once the pipeline was
installed and certified.
   Funding for the seepage control system, monitoring of the In-
itial Remedial Measures (IRM), the  Remedial Investigation (RI)
and the Feasibility Study (FS) was the sole responsibility of Land-
fill Inc. Upon completion of these activities, LI was eligible for a
50% reimbursement for these costs. The reimbursement was to be
drawn from the "sinking fund." If no remedial actions were re-
quired, LI was eligible for 100% reimbursement for the RI/FS sub-
ject to fund availability.  If remedial actions were required, LI could
perform these actions at its own cost,  subject to later reimburse-
ment as monies became  available in the "sinking fund." This
would also make LI eligible for 100% reimbursement for the RI/
FS. Alternatively,  should LI decide not to perform the remedial
actions, this responsibility would fall to Boulder County. LI would
then be eligible for only a 50% reimbursement for the RI/FS costs
until the  remedial actions were completed and "sinking fund"
money again became available.
INITIAL REMEDIAL MEASURES
   Within one year of the signing of the Cooperative Agreement,
the raw water pipeline had been designed, installed and tested.
This work  was supervised by the County, Public Works Depart-
ment,  City of Louisville Public Works Department and FRICO.
Final certification has  been delayed due to minor deficiencies re-
lated to valving and right of way.
   An integrated surface water management system was designed,
and construction of this system was also initiated. Although a con-
ceptual seepage control system was presented with the surface
water management system, construction of this system has been de-
ferred. This was done  to allow  integration of the seepage control
system with any remedial actions that may be proposed under the
FS.  Monitoring of the effectiveness of pipeline  installation has
been achieved by pressure testing the pipe and monitoring the flow
and chemical quality of the underdrain discharge.
REMEDIAL INVESTIGATION
  The Remedial  Investigation  consisted of three distinct tasks
performed within a phased approach. This work was supervised
by  a study group composed of representatives of the Boulder
County Health Department, CDH, LI and Fox Consultants. The
USEPA maintained indirect involvement with this group through
an  advisory role  to CDH. A phased investigation  was adopted
where the uppermost saturated  zone was characterized first; suc-
cessively deeper saturated zones  would be characterized only if the
overlying zone was contaminated.
  Task  1  required a literature  review of all previously collected
geologic, monitoring and operational data.  This task  was com-
pleted prior to the signing of  the  Cooperative Agreement and
served as a basis for scoping all of the investigative  activities and
for design of permanent monitoring program for the entire area.
  Task 2 consisted of all field investigations and laboratory analy-
ses. Under this task, 32 new piezometers and monitor wells were
installed; aerial photography was obtained and a topographic base
map and well survey network were developed; and 70 new and ex-
isting wells,  piezometers  and methane probes  were examined,
developed, stabilized and  monitored for field parameters includ-
ing temperature,  pH, conductance,  dissolved oxygen, explosive
gases and volatile organics  (using a photo-ionization detector);
water levels were measured on  a monthly or bi-monthly  basis;
and 23 water and leachate samples were collected and submitted
for full priority pollutant analysis.
  All of these activities  were completed, and a basic data report
describing the results of these activities and the procedures used
was developed and distributed within five months of the signing of
the Cooperative Agreement.
  Task 3 consisted of interpretive efforts related to site character-
ization and contaminant generation and migration assessment. A
report summarizing the geology,  hydrogeology, refuse conditions,
hydraulic structures, water quality and contamination characteriza-
tion was prepared and distributed seven months after the signing of
the Cooperative Agreement. Significant conclusions contained in
this report included the following:
•There appears to be a minimum of 50 to 60 ft of shale and minor
 interbedded sandstone between  the alluvial aquifer and the Lar-
 amie-Fox Hills aquifer beneath the site.
•Where piezometers have been completed in all four water bearing
 horizons, the potentionmetric surface of the uppermost portion
 of the  regional L-F aquifer, the "B"  sand is above the alluvial
 water table and above the ground surface.
•The major water bearing units  are separated from one another
 by low permeability aquitards which are a minimum of 20 ft thick.
•Approximately 15 to 25 ft of alluvium remains below the active
 landfill; however, only 0 to 5 ft remain beneath the inactive land-
 fill. As a result, saturated refuse can be found throughout the in-
 active landfill, particularly south of Community Ditch and in a
 small area in the southernmost portion of the active landfill.
•For much of its length, the base of the French drain is above the
 water table; however, along  its northern portion it appears  to
 collect both contaminated and uncontaminated water.
•Both Lagoon No. 1 and Lagoon No. 2 discharge some overflow
 into Community Ditch. This was solved by installation of the  60
 in. pipeline.
•There was no detected impact on the Marshall Lake-Community
 Ditch System as a result of contamination at the landfill. Con-
 tamination of Cowdrey Reservoir No. 2 appears to be limited to
 elevated concentrations of total dissolved solids and major ions.
•Groundwater contamination is characterized by: elevated concen-
 trations of total dissolved solids and major ions; increased levels
 of trace metals  including barium,  iron, and manganese; and
 volatile organics, principally chlorinated hydrocarbons and mono-
 cyclic aromatics.
•Volatile organic contamination  off-site occurs at concentrations
 an order of magnitude below those found on-site.
•The principal pathway for potential off-site migration of con-
 tamination is within the alluvial  aquifer along the east side of the
 inactive landfill and along Cowdrey Drainage.
  Subsequently, a second phase of investigation was conducted
during the summer of 1984 to  answer questions related  to the
shallow bedrock aquifer, the northern portion of the inactive land-
fill, the septic pumpage disposal ponds and  off-site contaminant
migration within the alluvial aquifer. At this time, only prelim-
inary results are available. Several conclusions may be drawn.
•The shallow bedrock  aquifer and deeper  bedrock  aquifers are
 generally isolated from  the  contamination within the  alluvial
 aquifer  by upward flow gradients and/or low permeability (10 ~5
 to 10-? cm/sec) clay shales.
•Contamination originating  in the  septic  disposal area and the
 northern portion of the inactive landfill displayed organic con-
 taminant concentrations one order of magnitude  higher  than
 those observed elsewhere in the area.
•Off-site contamination appears to be limited  to the immediate
 vicinity of the landfill (i.e., within 100 to 200 ft. of the boundary)
                                                                                            LEACHATE CONTROL
                                                                                                                           105

-------
and there is no evidence of a major contaminant plume off-site
within the alluvial aquifer.
FEASIBILITY STUDY

  Work on the FS will begin in September,  1984, once amended
basic data and contaminant assessment reports are completed. At
this  time three major types of remedial actions are being consid-
ered: (1) surface  controls, primarily surface water diversion,  re-
grading and upgrading of the  existing landfill cover; (2) ground-
water source controls, principally  interceptor wells/trenches and/
or impermeable barriers; and (3) leachate collection and treatment
using air stripping and/or carbon absorption.
 CONCLUSIONS

   The Cooperative  Agreement offered a  method  whereby in-
 volved, concerned and impacted parties could have direct partic-
 ipation in remedying a  longstanding environmental  problem.
 Through the Agreement, the Colorado Department of Health was
 able to initiate and be closely involved in the investigation and mit-
 igative actions at the State's  top priority  Superfund site. This
 agreement was  consummated  prior  to  any formal  legislation
 authorizing Colorado to participate in Superfund and at a time
 when it appeared that the legislature might not authorize the State's
 10% contribution to the fund.
   Boulder County took a lead role in resolving a  longstanding
 major environmental problem  in what was otherwise known as a
 progressive and "environmentally conscious" community. In addi-
tion, the Cooperative Agreement afforded them a means of "grace-
fully" supporting the much needed expansion of the Boulder Land-
fill at a time when the community perceived it to be a major en-
vironmental threat.
  The City of Louisville and FRICO have experienced major bene-
fits from the installation of the pipeline. These include not only the
protection of their water supplies, but also integration of the pipe-
line with existing capital  improvements projects. Louisville was
able to incorporate the pipeline into its overall program of expand-
ing its water supply and water treatment system. FRICO was able
to integrate the pipeline into program  for upgrading the Marshall
Dam outlet works and a long-term program of concrete lining the
entire Community Ditch system.
  Landfill  Inc.  has received benefits from its involvement in the
Cooperative Agreement. First, it  successfully obtained a permit to
expand its facility, thus avoiding  the loss of the capital investment
in the expansion plan and permit applications. Second, direct in-
volvement  and  control in the investigation provided a means  of
control on  the activities and expenditures. Third, indirect associa-
tion with a bad site and related potential negative public percep-
tion, which may have affected other operations in the state, was
averted  and  converted into  positive  public opinion and  local
government support.
  In summary, a group of parties not directly responsible for prob-
lems at the landfill but aware of potential negative impacts result-
ing from the problems at the site, was able to fashion an agree-
ment  to remedy this  problem that  was mutually  acceptable and
beneficial to all. Although modeled after the NCP, this Agreement
was developed and implemented without direct involvement by the
USEPA.
106
         LEACHATE CONTROL

-------
INVESTIGATION  AND  CORRECTIVE  ACTION: HOW  IT WAS
          DONE AT A  SUPERFUND SITE IN CONNECTICUT
                                               JAMES P.  MACK
                                           RICHARD C. DORRLER
                                            Fred C. Hart Associates
                                            Water Resources Division
                                              New York, New York
    INTRODUCTION
      The degree of risk or threat posed by any landfill to adjacent
    populations is highly dependent on site-specific conditions such as
    hydrogeology,  materials  disposed and local resource demands.
    While a general understanding of potential risks can be obtained
    from a preliminary analysis, often this determination may be mis-
    leading without site-specific information. This was the case with the
    Laurel Park Landfill in Connecticut.
      Initial evaluations, based on assumed hydrogeologic conditions,
    had predicted  significant impacts to downgradient private water
    supply wells. However, a detailed and comprehensive site investiga-
    tion produced evidence indicating that the initial model did not
    represent the site, and that potential impacts associated with leach-
    ate migration from the Laurel Park Landfill could be quickly and
    efficiently controlled by the installation of a perimeter toe drain.
      Of particular interest in this investigation was the bedrock eval-
    uation program. Early in the investigation, it was discovered that
    an understanding of the bedrock  aquifer was key to a determina-
    tion  of leachate migration pathways  and  potential impacts to
    groundwater. Toward this end, Hart Associates initiated a compre-
    hensive program design to evaluate bedrock fracture system, de-
    gree of weathering, presence  of leachate  in fractures and migra-
    tion pathways  available in the bedrock aquifer. The program con-
    sisted of test borings, rock cores  and the construction of two ex-
    tensive trenches on the east and south sides of the landfill.

    BACKGROUND
      The Laurel  Park Landfill is located at the top of Huntington
    Hill near Naugatuck,  Connecticut  (Fig.  1).  The site consists of
    about 30 acres, of which 19 are covered by fill. The property has
    been the site of waste management activities since  the late 1930s.
    Until the mid-1960s, the site received a combination of industrial
    and municipal  wastes. Since then, the site has received only munic-
    ipal wastes. West of the site  is Naugatuck State Forest.  North,
    east and south of the site are approximately 40 homes served by in-
    dividual private wells.

    FIELD INVESTIGATIONS
      In this section, the authors briefly describe the variety of field in-
    vestigations undertaken at the Laurel Park Landfill between Nov.
    23,1982 and Jan.  18, 1983. While the original intent of the investi-
    gation was to evaluate the effectiveness of a glacial till layer in pre-
    venting vertical migration, it was discovered early in the investiga-
    tion that the till was not as extensive as thought and consequently
    the goals of the investigation shifted to defining the hydrogeologic
    characteristics  of the bedrock and hydrodynamics of the landfill.
Test Boring and Well Installation Program
  A test boring program was conducted to define the thickness of
the landfill, identify the types and characteristics of the materials
underlying the landfill, locate the bedrock surface and  generate
subsurface sections. Four test borings were drilled through the
landfill, using a combination of hollow stem augers and mud rotary
(Fig. 2). Boring FTW-1 was intended to pass through the land-
fill, through the till and sample the bedrock surface. This process
required driving casing through the landfill materials, installing a
bentonite  seal, then inserting a smaller diameter casing through the
remaining soil materials. This double casing technique was de-
signed to  prevent landfill material from reaching  the bedrock. A 5
                         Figure 1
            Regional Location of Laurel Park Landfill
                                                                                         LEACHATE CONTROL
                                                      107

-------
ft rock core was obtained to observe the physical condition of the
bedrock.
   In addition to the four borings on the landfill, another boring
(TB-1) was drilled near the western  base of the fill. This boring
went through an  extensive fill layer and  was finished  as a  well
screened at the till/bedrock interface.  A 5 ft  rock core was ob-
tained in this boring to verify that bedrock, rather than a boulder,
had been reached.
   Three of the fill borings were finished as wells (Fig. 2). The  pur-
pose of installing  wells within landfill materials was to provide in-
formation pertaining to leachate behavior and character (i.e., con-
tamination source).  This  process included determining   the
saturated thickness of the leachate in the landfill, the direction and
gradient of leachate movement and the fluctuation  of leachate
levels in response to precipitation. These wells were also used to
sample leachate for chemical characteristics.
   Typical fill well construction is shown in Figure 3.  Before install-
ing any landfill  wells, a layer of bentonite pellets was placed in the
bottom of the borehole which extended several feet above the bed-
rock/till interface.  This was done to prevent  leachate migration
through the bore hole. Wells consisted of 2 in. diameter flush joint
 PVC screens and  casing which were installed through the augers.
 Sand or gravel packing was placed in the annular space around the
well casing and brought to within 6 to 10 ft of  the surface. A ben-
tonite seal was placed on top of the sand pack, and the remaining
 annular space was  filled with cement  grout. A protective casing
 with locking cap was added as a final step. Water level monitoring
 of these wells showed fluctuations in leachate levels, indicating that
 these wells are  in communication with the landfill  and  are func-
 tional observation wells.

Test Pits and Trenches

   Seven test pits and two trenches were  constructed to gain further
understanding of  the character and extent of the till and  to deter-
mine the conditions of the bedrock. Each test pit  was  dug  to a
depth of 10 ft with  a tractor-mounted backhoe. Samples were ob-
tained from the side walls,  and the pits were logged by the super-
'•IT
• CLOW
QHOUHC
0-
10-.
30-
40-
• 0-
•
to -
ro-
• 0-
• 0-
;
100 -
-1
B c
n °
~






X
in
Ł




v
- 4' PROTECTIVE
STEEL CAS»«
^r.o««u

^-rrvc CASMO


^^TMCAOEO PVC JOMT

^^ »J' 10 SLOT PVC SCREE*




ii«'« WELL DIAGRAM
FOR WELL FW-4

         Figure 3
Well Diagram for Well FW-4
                        LAUREL PARK LANDFILL
                      LOCATION Of TEST BORINGS
                   A f <• * ret! Banna cammo AS WILL
                        OCPTH Of
                                                            Figure 2
                                                       Laurel Park Landfill
                                                     Location of Test Borings
108
         LEACHATE CONTROL

-------
vising geologist. A shallow well was installed in each pit before
backfilling. Soil encountered in all pits was very consistent, con-
sisting largely of gray silty and clayey medium to fine sand. The
till soils were found to be very dense. Bulk samples were obtained
from three test pits and sent to a soils laboratory for permeability
analysis. Permeability ranged from6.8x 10~5 to 1.5 x 10~6 cm/sec.
  Trenches were used because they are an effective means of pro-
viding a continuous horizontal exposure of soil  or rock  along a
given line or cross section. One trench was excavated along the en-
tire southern side  of the landfill  with  a bulldozer,  and  another
trench was excavated along the entire eastern side with a backhoe.
The south side trench showed that bedrock was within three to four
feet of the surface and that the bedrock exposed was relatively  un-
weathered Straits Schist. The east side trench was 1400 ft long and
also  exposed  Straits  Schist with  unusually  consistent foliation
orientation.

Surface Runoff Measurements

  The drainage conditions at the Laurel Park Landfill were such
that  a major portion of site runoff discharged to  a small creek on
the north side (Fig. 4). In order to estimate the various components
of flow contributing to the runoff volume, this stream (which orig-
inated at the landfill) was chosen for a measuring location. An 18
in. diameter corrugated pipe was installed at the streams discharge
point from the site and an earthern dam was constructed so that all
surface runoff from the landfill exited through the pipe. Measure-
ments of flow were made by placing a 5 gal pail beneath the pipe
and recording the amount of time required to fill it.
  The information developed clearly indicates that the vast major-
ity of precipitation that falls  on the site exits as surface runoff
rather than infiltrating to form leachate. The discharge measure-
ments at the culvert (from Nov., 1982 to Jan., 1983) closely corre-
lated with rainfall events, indicating the immediate response of sur-
face runoff to precipitation. During periods of no precipitation, the
average culvert flow  of 3.2 gal/min consisted predominantly of
leachate.

Seep Mapping and OVA Field Screening
  During  the   field  investigation,  major leachate  seeps were
mapped. The information, in conjunction with data on leachate
evaluations in wells,  provided further  understanding of  the  hy-
draulic properties of the landfill. Only leachate seeps with  obvious
discharge points and definitive evidence of past or  present flow
were mapped.  The vast majority  of leachate seeps  were located
at or near the base of the fill. The most active seeps  were located
near the northern base of the  fill.  These formed a horizontal line
approximately  3 ft above the  base of the fill and flowed contin-
uously. Seeps along the  east side consisted of isolated outbreaks
and  flowed intermittently. When  leachate seep  elevations were
correlated with leachate elevations in the wells, it was determined
that  the leachate body in the landfill sloped northward similar to.
the bedrock surface base of the landfill and was discharging at  the
base of the northface.
  Field screening with  an Organic  Vapor Analyzer  (OVA) was
also conducted for the primary purpose of detecting leachate seeps
downgradient from the landfill. A perimeter survey was conducted
to measure total airborn hydrocarbons. There were  no leachate
seeps detected anywhere beyond the immediate base of the landfill.

Electrical Resistivity Survey
  Two electrical resistivity survey runs were conducted along  the
east  side of the landfill. These surveys used the horizontal pro-
filing approach in which a constant electrode spacing of 40 ft was
maintained. It was thought that by  comparing resistivity values  for
unknown  areas  to values obtained for an area known to be un-
contaminated, areas of possible contamination could be identified.
However,  many factors  affected the resistivity values,  including
depth to bedrock, amount of moisture present in the  soil and  the
characteristics of the saturating fluid. Therefore, electrical resistiv-
          4JS-



          400-



          37S-



          350-
           50-


           25-
I   I
                               DIRECT

                              SURFACE

                               RUNOFF
                               PRECIPITATION
      CO  090-
             NOV
                          DEC
                                        JAN
                           Figure 4
       Relationship Between Surface Runoff and Precipitation
ity surveys were used merely as a field screening technique and
were followed by more detailed investigations.
  For the first run, apparent resistivity values ranged between 800
and 4,000 ohm-ft. Because of the variation between shallow bed-
rock and moist soil conditions, interpretation was difficult. Low
values measured in an area of surface runoff from the landfill may
be due  to  shallow  soil  contamination. The tentative conclusion
reached using the resistivity survey was that runoff from the land-
fill is effecting shallow soil in certain areas east of the landfill.
                                                                                             LEACHATE CONTROL
                                                                                                                            109

-------
Bedrock Evaluation Program
  Through the test boring program,  it was discovered that por-
tions of the east, northern and southern sections of the landfill were
either resting on a thin till soil layer covering the bedrock or were
directly on the bedrock. Thus it was decided to further evaluate
the characteristics and configuration of the bedrock surface in the
immediate vicinity of the landfill, as the condition of this surface
would be critical to the assessment of the potential for contam-
inant migration from the landfill through the bedrock aquifer.
  In a  bedrock aquifer, the principal conduits for groundwater
flow are the joints and fractures. It was reasoned that if the Laurel
Park Landfill is resting  on the bedrock surface, and if leachate is
entering the bedrock, then cracks  or  joints immediately adjacent
to the landfill edge would be the first  to show signs of contamina-
tion.  The evaluation  program consisted of constructing  two
trenches (Fig. 5) and then carefully cleaning the rock at select loca-
tions (using shovels, brooms and high pressure water jets) to ex-
pose the worst case fracture conditions for observation, measure-

   The rock type consisted predominantly  of a hard, grey, coarse-
grained mica schist (Straits Schist). The rock surface ranged from
hard and smooth  to  rough,  irregular  and  blocky.  Chemical
weathering  plus  freeze  wedging,  root activity and glacial pluck-
ing have been the dominant processes acting to open the fractures.
Fracture widths ranged  from 0.25  to 6 in., but in every case these
openings  narrowed to hairline cracks  within a few feet of the sur-
face. Prior to cleaning,  these cracks were  filled with compact clay
soil. Thus, it was determined that shallow surface weathering pro-
cesses were the only plausible mechanism for producing open frac-
tures. There was no evidence of contaminant migration in any frac-
tures.
FINDINGS
   Based on the information obtained  from these various investiga-
tions,  the geologic and hydrogeologic characteristics of the site
were established.  The goal was  to determine the  leachate migra-
tion pathway and define  groundwater conditions  adjacent to the
fill. The information provides a basis for understanding potential
environmental impacts associated with the  flandfill.
                                                        Site Geology
                                                           The results of the investigations demonstrated that the
                                                        underlying the site is part of the resistant Straits Schist, as described
                                                        by Carr.' The orientation of the vast majority of fractures in the
                                                        surface of the exposed bedrock are predominantly parallel to the
                                                        foliation. Fractures  did not show evidence of displacement, indi-
                                                        cating that faults or shear zones are not present. Fracture widths
                                                        at the surface were as much as 6 in., but quickly tapered to hair-
                                                        line cracks. This is consistent with observations made by Ellis,' who
                                                        found that in artificial cuts, such as a quarry, joints that may be
                                                        open 0.5 in. at the surface were found to be very tight at 25 ft below
                                                        the surface.  Rock coring near the landfill also confirmed this ob-
                                                        servation.
                                                           Weathering of  the  exposed  bedrock  surfaces was  largely de-
                                                        pendent on the minerology and degree of fracturing. In areas where
                                                        the bedrock  was predominantly composed of feldspar  and quaru
                                                        crystals, there was little physical and chemical weathering; in ex-
                                                        posures where the rock was fractured and composed of muscovite
                                                        and biotite, weathering was more extensive. However, though root
                                                        penetration and frost  wedging  had enlarged  fractures  at  the sur-
                                                        face, they  quickly narrowed, indicating that the effect of weather-
                                                        ing is limited to the upper 5 or 10 ft of the rock.
                                                           As shown  in Figures 6 and 7, the bedrock surface forms an elon-
                                                        gated ridge orientated  in a north-south direction under  the eastern
                                                        portion of the landfill. Bedrock is at or near the surface to the east
                                                        of the fill, and till underlies the western portion.
                                                           A wedge-shaped body of low  permeability glacial till overlies the
                                                        bedrock in the western portion of the site (Fig. 7). The  till consists
                                                        of a dense, gray, clayey, fine to medium sand with trace fine gravel
                                                        to coarse  sand. Gneissic and schistose  boulders are  scattered
                                                        throughout the unit. Test  pit and  test  borings indicate  that till
                                                        textural composition is relatively consistent with depth.

                                                        Site Hydrogeology
                                                           The  site consists of three distinct water bearing units; the bed-
                                                        rock  system, the till and the landfill body. Water in the bedrock is
                                                        confined to secondary openings such as joints or other fractures.
                                                        Therefore, fractures within the rock become the predominant stor-


.•
•
LAUREL
>
LOCATION OF f
••*• MOHOCK i
•n-t Tirr MMM
nc MMOC
• ITIIKI AMI
PARK LANDFILL
JEDftOCX EXPO3LFCS
IPO BUM
0 H WMCM A COM OP
Jl WAB O«TA»aO
0» OP MOHOC« OUTCMV
                                                       //
                 .1» •FAUCI
                 /   mooM
                                                            Figure 5
                                                  Location of Bedrock Exposures
 110
LEACHATE CONTROL

-------
age area for water, and the continuity, frequency, width and degree
of interconnection among the fractures dictate the overall water
bearing properties of the rock mass. As mentioned previously, field
observation and previous studies2'3 support the authors' contention
that the width and interconnectness of fractures decrease rapidly
with depth.
  Schist, in particular, is susceptible to fracture closure. Because it
is less competent than other metamorphic rock, schist responds to
regional stresses by minute  bending or  slipping along  foliation
planes, similar to a relaxation that is distributed internally through-
out the rock mass. Although fractures develop  in schist, they are
likely to be nearly closed and discontinuous and, therefore, poor
conduits for water.3 Also, the physical and chemical mechanisms
for widening fractures are all concentrated at the surface and have
little effect on the width of fractures 10 ft or more below the sur-
face.
  Based on literature research, it was concluded that if any inter-
connected fractures existed in the bedrock, they would  be within
100 ft of the surface; between 100 and 200 ft there may  be a few
thin fractures, but these are limited in extent and interconnectivity.
Below 200 ft, fractures would be essentially nonexistent. Based on
the intense investigation of the bedrock within the immediate vicin-
ity of the landfill, it was concluded that the vast majority of water
in the bedrock actually flows  within 10 ft of the surface.
   Because flow in the bedrock is restricted to the upper few feet,
topography and bedrock  surface configuration have the  dominant
control on groundwater flow direction. Thus, the shallow ground-
water flow  divides would be expected to  closely approximate sur-
face  water  drainage divides and bedrock highs. A groundwater
divide underlies the landfill slightly east of well FW-4. East of the
divide,  due to the very steep topography, groundwater flow is dis-
charged to the surface or moves at  the soil/bedrock  interface.
West of the divides, in addition to topographic controls, ground-
water flow  is also influenced by the northwest-southeast fracture
orientation. The combined effect is to direct  groundwater flow to
discharge to the creek at the north end of the site.
   The till is a compact poorly sorted material of fine-grained clay
size particles resulting  in a  low  permeability. Laboratory  perme-
ability measurements from bulk samples taken at the site  ranged
from 6.8 x  10"6 to 1.5  x  1~5 cm/sec and  overaged 4.0 x 10'' cm/
sec. Calculations indicate that lateral seepage, using a permeability
of 10~6 cm/sec, porosity of 10% and a hydraulic gradient of 0.1,
would be approximately 1 ft/yr. This value of the hydraulic grad-
ient indicates that water held in the till is subject to very little lateral
or vertical movement and that the till can be considered as a sat-
urated body, rather than an aquifer.
  Although not a natural geologic deposit, the landfill can be con-
sidered as part of the site hydrogeology because it is in direct com-
munication  with the underlying glacial till and bedrock and con-
tains a saturated zone with definite lateral flow. However, the evi-
dence indicates that the leachate,  rather than moving downward,
is flowing horizontally and exiting the landfill in  seeps around the
base. The leachate level in the fill is unusually low when compared
to fill relief, indicating thorough lateral drainage of the landfill
body.
  In summation, it has been determined  from this investigation
that, while the landfill is in direct communication  with the underly-
ing till and bedrock, practically no leachate is entering the under-
lying geologic materials. This is because the underlying formations
have very low permeabilities and,  as such, water held in the frac-
tures of the bedrock or the interstices of the till is subject to very
limited lateral or verticle movement. In essence, these two units can
be considered as saturated but with very little groundwater  move-
ment, and they are acting as a natural liner. Leachate in the land-
fill flows over this stagnant water layer and exits out the base of the
fill.
  The leachate layer mounds in an arch similar to the curve of the
bedrock surface under the fill, which results on a high point  slight-
ly east of well FW-4. This causes a driving head  that parallels the
bedrock or till surface, forcing the leachate to flow along the bed-
rock/fill or till/fill interface.  The leachate level  is parallel  to the
till/fill interface and several seeps on the north face of the landfill
connect with this level. The seeps at this end of  the landfill were
observed to have a continuous discharge, indicating that this is a
preferred flow direction for the leachate.
Groundwater and Leachate Quality

  Between June and Nov., 1982, residential supply wells surround-
ing the site  were sampled by the Conn.  Dept. of Health Services
(DOHS). Additionally, on Jan. 18 and 19, 1983, the leachate from
                                                                                 To LOHO MEADOW
                         LAUREL PARK LANDFILL

                              •OAlflFItT)
                      BEDROCK SURFACE CONTOUR MAP
                       •rf CONTOUR UC (CONTOUR tfTERVAL • 10 FT)
                        *  ROSS HECTWN
                     . M.  KDROCK BLfFAOE ELEVATION
                     * **T  BY T18T Kf*JQ
                      X   BEDROCK SURFACE
                     m  ELEVATION iCAOLFCO H TRENCH
                     . u.  BEDROCK BUT ACE ELEVATION DETERMrfl)
                         ATOurcnoPt
                                                              Figure 6
                                                     Bedrock Surface Contour Map
                                                                                               LEACHATE CONTROL
                                                                                                                               111

-------
                                                                          tuBauprt.ee CBOM atcrtOM  A-A'
                                                                          LAUREL PAW LA/CfU.
                                                                                                                A'

                                                                                                              IA«T
                                                             Figure 7
                                          Subsurface Cross Section A-A' Laurel Park Landfill
                                          LEACHATE MOUNO
                DOWNWARD POCaeURC OHADICKT
                  mam. TWO m MIGRATION TO
                       QftOUMOWATCR
ll_ACtAL TIU-
                                           LEACHATE LEVEL
                                              LfACHATC MIGRATION
                                                IMPACT* MJRPACC
                                                    WATCH
                          BEDROCK
                           Figure 8
       Comparison of Typical Landfill and Laurel Park Landfill
 three landfill wells  (FW-2,  FW-3 and  FW-4)  was sampled and
 analyzed. The results of the leachate analysis showed a wide variety
 of volatile organic compounds, some at fairly high concentrations.
 If leachate from the landfill was migrating through the bedrock
 awuifer to these wells, then it is logical to assume that after 20 yr of
 landfill operation and potential migration,  that some of the com-
 pounds would have been detected in the residential wells. However,
 of the 36  wells samples by  DOHS, only one  well showed com-
 pounds found in the leachate and this only at trace levels.
   In order to fully evaluate the potential for migration, a complete
 groundwater monitoring system  is presently being installed at the
                                                                                                                         TMMCH 1-1
                                                                                                                       •v.
^REFUSE  , fOVER MATERIAL!*	  5'-
                xRECOWACTED TILL
                ' ">i-'"'0
                                         FILTER FABRIC ^SCHIST
                       Figure 9
                  I aurel Park Landfill
                 East Trench Toe Drain
 112      LEACHATE CpNTROL

-------
landfill. It will consist of wells at various depths on the east, south
and north sides. Rock core is also being obtained during this effort
to further evaluate the characteristics of the bedrock aquifer.
CONCLUSIONS

  Geologic and hydrogeologic information obtained from this site
investigation indicates that the eastern portion of the Laurel Park
Landfill lies directly on the bedrock surface or a thin till covering,
while the western portion is separated from the bedrock by a wedge
of low permeability glacial fill. This information, accompanied by
data revealing that the bedrock and till are at their maximum ele-
vations beneath the landfill  and slope downward to the east, west
and north, indicates that the most likely pathway of leachate mi-
gration will be in the horizontal direction, with leachate exiting the
fill at or near the base of the slope and draining off the mountain
via the surface water drainage.
  As shown by Figure 8, this situation is different than typical
landfill conditions, in which groundwater impacts from  leachate
migration can be expected due to a build-up of a leachate head in
the waste cell causing a downward pressure gradient that results in
eventual downward migration. If the underlying formations are sig-
nificantly permeable, then the leachate will enter the groundwater
system. However,  at the Laurel Park  Landfill,  the underlying
formations are very poorly  permeable and there is no deep com-
ponent of groundwater flow (400 to 500 ft) as  postulated by past
reports. Therefore, it is extremely unlikely that a pathway exists be-
tween  the landfill and the downgradient water supply wells, and
there is no conclusive evidence of groundwater contamination from
the landfill.
  Because the majority of leachate runs directly off the site into the
surface drainage system, an effective leachate collection system is
feasible. A network of toe drains and  rock trenches has been in-
stalled adjacent to the base of the landfill to collect the near surface
groundwater and leachate and convey it by gravity flow to the Nau-
gatuck Wastewater Treatment Plant for treatment and disposal.
  The leachate collection system consists of two separate systems,
an East and a  West Trench Drain. The East Trench Drain  con-
sists of approximately 1,750 ft of 4 in. diameter ABS pipe placed
within a gravel envelope (Fig. 9). The  perforated  leachate collec-
tion pipe and gravel envelope have been installed in a trench 10 ft2
in cross-section and cut through rock on the south and east sides
of Huntington Hill. The West Toe Drain consists of approximate-
ly 2,400 ft of 4 in. diameter ABS pipe placed in a  gravel envelope
in the till along the south, west and north sides of the landfill. The
installed drain system should provide a  highly effective mechanism
controlling leachate excape from the landfill.


REFERENCES

1. Carr, M.H.,  The Bedrock Geology of the Naugatuck Quadrangle,
   State of Connecticut Geological and Natural History Survey, Quad-
   rangle Report No. 9, 1960.
2. Ellis, E.E., "A Study of the Occurrence of Water in Crystalline Rocks",
   in Gregory, H.E. (editor) Underground  Water Resources of Connecti-
   cut, U.S. Geological Survey Water Paper 232, 54-103.
3. Wilson, W.E., Burke, E.L. and Thomas, C.E., Water Resources In-
   ventory of Connecticut, Part 5, the Lower Housatonic River Basin,
   Connecticut Water Resources Bulletin No. 19, 1974.
                                                                                              LEACHATE CONTROL
                                                           113

-------
    EVALUATION  OF  "SUPERFUND"  SITES FOR  CONTROL
          OF LEACHATE  AND CONTAMINANT MIGRATION

                                             ROBERT S. McLEOD
                                            Engineering-Science, Inc.
                                                 Atlanta, Georgia
INTRODUCTION
  Engineering-Science, Incorporated has been involved in studies
relating to the management of hazardous wastes at sites throughout
the United  States for a  number of years.  Assessment studies,
remedial alternative studies, treatment process studies and im-
plementation of remedial action plans have been included as a part
of these hazardous waste management studies.
  The purpose of this paper is  to share some experiences that per-
sonnel  at Engineering-Science have had  in studies  relating to
leachate control  and control of  migrating groundwater that has
been contaminated.

LEACHATE CONTROL AT
TAYLOR ROAD LANDFILL
  Taylor Road Landfill is an unlined 42.5-acre municipally owned
landfill that was operated by Hillsborough County, Florida from
1976 to 1980. The landfill received approximately 800 tons/day of
municipal solid waste during its operational life.' The landfill was
closed in February, 1980 and a cover consisting generally of clayey
sands and sand was placed over the solid waste.
  The USEPA sampled groundwater at and near the landfill in Oc-
tober, 1979. Volatile organic compounds were present in the water
around the landfill. A subsequent study' sponsored by the County
confirmed the presence of contaminants in the  groundwater and
outlined the approximate extent of the contamination.
  Mechanisms whereby the landfill would contaminate the ground-
water include:
•Leachate produced by the groundwater  levels rising  above the
 base of the landfill
•Leachate produced by rainwater percolating downward through
 the landfill
•Liquid waste discharging from  containers buried in  the landfill
One or more of  these mechanisms could be contributing to con-
tamination of groundwater in the area of the landfill.
  This  study began in November,  1981 with a geophysical  in-
vestigation to help define the  base of the landfill. The primary
mechanism for groundwater contamination was initially believed to
result from groundwater contacting the base  of the landfill during
periods of high groundwater levels.
  The results of the geophysical investigation,  together with an
evaluation of groundwater level trends in the area  of the landfill,
indicated that the base of the landfill was probably  above the
general water table. However,  it  was determined that periodically
perched water in the area of the landfill could  contribute to the
generation of leachate at the landfill.
  The  condition of  the  landfill  cover  observed  during  the
geophysical investigation suggested that the cover might not be ade-
quate and that downward percolation of surface water might be a
significant  contributor  to the production  of  leachate. Standing
water was observed in placed on the landfill cover and in the sur-
rounding drainage ditches, and large cracks were observed in the
cover in other places.
  In October,  1982, a subsequent  study  to evaluate the effec-
tiveness of the landfill cover was conducted.
Purpose and  Scope
  The purpose of this study was to determine  how effectively the
existing landfill cover and drainage prevented percolation of water
to the waste materials.
  The scope  of work included a field  investigation at the Taylor
Road landfill, computing a water balance for the cover overlying
the landfill and evaluating the probable hydrologic impacts on per-
colation of upgrading the cover and surrounding drainage.

Site Description

  The landfill is located in eastern Hillsborough County,  Florida
and is approximately 10 miles east of Tampa (Fig. 1).
  The landfill is situated on  top of a  north-south  trending ridge
that ranges between elevations of 80 ft and 120 ft above mean sea
level. This ridge probably reflects a high  area in the underlying
limestone bedrock.
  The regional topography is karst in nature. Numerous sinkholes
can be identified near the landfill study on the USGS Thonotosassa
and Brandon topographic maps.
Field Data Collection Program
  The  objective of this investigation  was  to define  the physical
characteristics of the cover overlying the Taylor Road  landfill. The
field data-collection program  included infiltration tests of the land-
fill cover, determination of cover thickness  and sampling the cover
and drainage ditches for laboratory analyses of selected soil proper-
ties.
  Sixteen infiltration  tests were conducted on the Taylor Road
landfill. The test sites were positioned in a rectangular pattern to
obtain a uniform distribution of observations over the  area (Fig. 2).
  The  infiltration tests were conducted in general conformance
with the American Society for Testing  Materials (ASTM) standard
test  for measuring  infiltration rate of soils in the field  using a
double-ring infiltrometer.' Two cylinders,  one inside the other,
were driven into the ground. The outer 36-in. diameter cylinder was
driven 6 in. deep. The inner 12-in. diameter cylinder  was driven 2
114
         LEACHATE CONTROL

-------
                            Taylor Road Landfill
                       Figure 1
                 Location of Study Area
                                                -i tfai
•AM MOOV» now MU.MOIKXMH COUMTY ACTIAI. fMOTO«A«<. lilt



    •XPtAMATIOH

4.1  • WflLTHATION T1IT AMD IOH. lAyrima ilTM

0»  . lOH. UUPIM ilTt WIY                 1	1	L_
                         Figure 2
               Location of Hydrologic Data Sites
in. deep. A constant water depth of 6 in. was maintained within the
cylinders during testing by adding water at regular intervals. Thir-
teen infiltration tests were 8 hr in length. Three 24-hr tests were
conducted.
  The vertical hydraulic conductivity of the landfill cover at its sur-
face varies from about  5  x 10 ~5 cm/sec to less than 5 x  10 ~6
cm/sec. This conductivity range is derived from the 8-hr infiltration
tests by assuming that the steady-state infiltration rate at a site, or
vertical hydraulic conductivity,  will be one-half the measured 8-hr
rate. This assumption is reasonable based on the results of the 24-hr
tests that were  conducted  coupled with the trends observed in in-
filtration rates during the tests.
  Soil samples from selected depths at each infiltration test site and
from test sites along drainage ditches adjacent to the landfill were
subjected to laboratory  analysis of vertical  hydraulic conductivity
and grain size. The samples for  testing were collected using Shelby
tubes. Vertical hydraulic conductivity was measured using a con-
stant head permeameter.  Grain size analyses were conducted in
general conformance  with  ASTM  standards.2'3  The vertical
hydraulic conductivities for soil  samples from  five  sites at the
Taylor Road landfill were too great to measure. The soils at these
sites consisted  mostly of  sands.  The vertical hydraulic conduc-
tivities for soil samples from the other  sites varied between 1.2 x
10~5 and 2.0  x 10-9 cm/sec. These samples consisted mostly of
clayey sand. The vertical hydraulic conductivities for soil samples
from six of the seven drainage ditch sites consisted mostly of sand
and were too great to measure with a constant head permeameter.
  The thickness and composition of the cover overlying the landfill
was determined at each of the infiltration sites by test borings.  The
maximum observed cover thickness at the landfill is greater than 48
in.  The minimum observed cover thickness at  the landfill is 8 in.
The cover is composed of clayey sands, sandy clays and sand.

Interpretation of Findings
  The probable hydrologic impacts of upgrading the landfill cover
and drainage ditches were  investigated. The effectiveness of reduc-
ing percolation through the landfill cover by increasing the average
cover thickness and decreasing the average cover vertical hydraulic
conductivity was studied.  Also,  the  impacts on percolation of
runoff to the water table by improving the drainage ditches at the
landfill were addressed.
  Average annual percolation was computed for the Taylor Road
landfill cover  based on a monthly soil  water balance. Average
monthly infiltration into the landfill cover was determined from a
surface water balance for the cover (Table 1). Infiltration was com-
puted as the difference between average monthly precipitation and
average monthly runoff. Average precipitation at the landfill for
each month was assumed to be equivalent to  average monthly
precipitation recorded at Plant City, Florida for the period 1951 to
1979.  Average monthly  runoff was computed  by computing
average annual runoff at  the landfill using the Soil Conservation
Service Curve Number method and distributing this runoff in ac-
cordance with the percentage of average annual rainfall that occurs
each month. The results of the infiltration tests on the landfill cover
were used to aid in these computations.
  Monthly  soil  moisture changes  in the cover  and monthly
evapotranspiration from the cover were computed using methods
outlined by the USEPA.4 Soil moisture changes  are  based on
monthly changes in infiltration and potential  evapotranspiration.
Monthly evapotranspiration was computed based on monthly in-
filtration rates, monthly  potential evapotranspiration  rates and
computed changes in monthly soil moisture.
  Percolation through the landfill cover is negligible for  average
rainfall (Table 2). Infiltration into the cover is not great enough to
overcome  evapotranspiration demands  and  deficiencies in soil
moisture storage.
  Percolation was computed for above normal rainfall conditions.
Yearly rainfall was assumed to be 140%  of  normal rainfall, or
74.92 in., for the simulated wet condition. Yearly rainfall totals in
excess of that simulated here should occur about one year in five on
                                                                                           LEACHATE CONTROL
                                                                                                                          115

-------
                          Table 1
         Surface Water Balance for Taylor Road Landfill
                Cover for Normal Precipitation
Month
January
February
March
April
May
June
July
August
September
October
November
December
TOTALS
Precipitation
(In.)
2.53
3.33
3.75
2.09
4.05
7.16
8.05
8.65
6.81
2.89
1.93
2.28
53.52
Runoff
(In.)
1.94
2.55
2.87
1.60
3.10
5.48
6.15
6.61
5.21
2.21
1.48
1.74
40.94
Infiltration
(In.)
0.59
0.78
0.88
0.49
0.95
1.68
1.90
2.04
1.60
0.68
0.45
0.54
12.58
                           Table 2
            Soil Water Balance for Taylor Road Landfill
                 Cover for Normal Precipitation
Month
January
February
March
April
May
June
July
August
September
October
November
December
TOTALS
Infiltration Soil Moisture Evapotrans-
(in.) Change plrallon Percolation
(In.) (in.) (In.)
1.94
2.55
2.87
1.60
3.10
5.48
6.15
6.61
5.21
2.21
1.48
1.74
40.94
0.61
1.16
0.41
-1.20
-0.71
-0.16
-0.12
-0.04
-0.04
-0.19
0.00
0.28
0
1.33
1.39
2.46
2.80
3.81
5.64
6.27
6.65
5.25
2.40
1.48
1.46
40.94
0
0
0
0
0
0
0
0
0
0
0
0
0
                          Table 3
           Soil Water Balance for Taylor Road Landfill
             Cover for Above Normal Precipitation
Soil Moisture Kvupotrani-

Month
January
February
March
April
May
June
July
August
September
October
November
December
TOTALS
Infiltration
(In.)
2.71
3.57
4.02
2.24
4.34
7.67
8.61
9.25
7.29
3.10
2.07
2.43
57.30
Change
(In.)
0
0
0
-1.22
-0.51
1.38
0.35
0
0
-0.74
-0.08
0.82
0
plrallon
(In.)
1.33
1.39
2.46
3.46
4.85
6.29
6.92
6.73
5.71
3.84
2.15
1.46
46.59
Percolation
(In.)
1.38
2.18
1.56
0
0
0
1.34
2.52
1.58
0
0
0.15
10.71
the average.7 Average monthly evapotranspiration conditions were
assumed for the computations.
  Annual percolation through the Taylor Road landfill coyer for
the simulated wet condition was 10.71 in. (Table 3). Percolation oc-
curred during  the winter months  and again during the summer
months. The corresponding average annual volume of water per-
colating through the cover to the  solid waste was approximately
12,000,000 gal.
  The hydrologic impact of increasing the average cover thickness
at the Taylor Road landfill by 24 in. was investigated for above nor-
mal precipitation conditions. The simulated annual rainfall was
74.92 in.
  Annual percolation through the  landfill cover for the simulated
wet condition  and added soil cover was 10.46 in.  This percolation
rate is not substantially different from the 10.71 in. of annual per-
colation computed without the added cover thickness. Increasing
the average cover thickness by 24 in. does not substantially reduce
percolation through the cover during wet periods.
  Decreasing  the average vertical hydraulic conductivity of the
landfill cover would substantially decrease the amount of infiltra-
tion into the cover and, hence, percolation to the solid waste during
wet periods (Table 4). The maximum amount of annual infiltration
to the solid waste would only be limited by the amount of rainfall
that occurred if the hydraulic conductivity of the landfill cover was
1  x  10-' cm/sec. The maximum annual infiltration would be
limited to approximately  12 in.  of rainfall  or 14,000,000 gal  by
 reducing the average hydraulic conductivity  of the landfill cover to
 1 x  10'° cm/sec and  1 in. or 1,000,000 gal by reducing the average
 hydraulic conductivity of the landfill cover to 1 x  10-* cm/sec and
 1 in. or 1,000,000 gal by reducing the average hydraulic conductivi-
 ty of the landfill cover to 1  x 10 - 7 cm/sec.
   It appears reasonable, based on the above analysis, that a vertical
 hydraulic conductivity of 1  x 10- ^ cm/sec or less for the landfill
 cover would substantially reduce or eliminate percolation through
 the cover. The  current  landfill  cover has an  average vertical
 hydraulic  conductivity somewhere between  1  x  10~5 and  1 x
 10-' cm/sec, which could potentially allow large  quantities of per-
 colation during wet periods.
  Percolation  of significant quantities of water to the water table in
the vicinity of the landfill  could  be eliminated  or significantly
reduced by improving surface drainage at the landfill.

                           Tabk 4
              Effect of SoU Cover Vertical Hydraulic
                   Conductivity  on Infiltration
Vertical
Hydraulic
Conductivity
(cm /sec)
10-5
10-'
10-'
Maximum
(in.)
124
\2A
1.2
Annual Infiltration
(million gals)
143
14.3
1.4
                                                                   The average annual volume of runoff that percolates to the water
                                                                 table  in ponded areas along  the east and south drainage ditches
                                                                 (Fig.  3) is equivalent to approximately  11,000,000 gal/yr or 21
                                                                 gal/min. Some of the runoff that flows to  the  north  and west
                                                                 drainage ditches also percolates to the water table. The material at
                                                                 the base of the drainage ditches is generally sand which would allow
                                                                 rapid percolation of water.
                                                                   Regrading the east and south  drainage ditches to eliminate the
                                                                 ponded areas  and to direct runoff away from the landfill would
                                                                 suppress most  of percolation that is currently occurring from these
                                                                 drainage ditches.  Lining all of the draining ditches with a clay or
                                                                 other relatively impermeable  material would  reduce or  eliminate
116
         LEACHATE CONTROL

-------
                                         8a°17'30"
                                                    500 FEET
         EXPLANATION


       - CENTERLINE ALONG DRAINAQE DITCH

         AREA WHERE WATER PON03

         SOLID WASTE CELL
                         Figure 3
     Relation Between Surface Drainage and Solid Waste Cells
percolation through  the bottom of the drainage ditches during
runoff events.
Remedial Actions
  Measures are currently being taken by Hillsborough County to
ungrade the landfill  cover and drainage ditches surrounding the
landfill. A minimum  24-in. thickness of cover is being placed over
the landfill in areas where the cover is less than 24 in. in thickness.
A minimum 6-in. thick soil layer whose permeability is less than 1
x 10 - 7 cm/sec is being placed over the cover in areas where testing
has indicated  that the permeability of the existing cover was inade-
quate. Also, the drainage ditches are being upgraded to reduce per-
colation of water, and ponding is being eliminated in the southeast
and south areas of the landfill.
CONTROL OF CONTAMINATED
GROUNDWATER MIGRATION

  A client is required to perform certain work at his plant site in
South Carolina as a result of a recent settlement reached between
the client and the South Carolina Department of Health and En-
vironmental Control  (DHEC).  Part of this  work involves im-
plementing a groundwater recovery  and  treatment program  for
contaminated groundwater.
Purpose and Scope
  The purpose of this study was to identify the extent of ground-
water contamination in the vicinity of the plant site and to design
and implement a groundwater recovery and treatment system.
  The scope of work included field data collection and laboratory
analysis. Field data collection efforts included a geophysical survey
to aid in defining the extent of groundwater contamination, an ex-
ploration drilling program to aid in determining the hydrogeology
of the area, base flow measurements along streams near the plant
site to define surface  water-groundwater relationships and  sam-
pling  water from wells  and  from  streams  for testing in the
laboratory. Groundwater  and surface water were analyzed for
selected chemical constituents  that  would be indicators of con-
tamination.
Field Data Collection Program
  The objective of the field data collection program was to collect
sufficient  data for developing a groundwater recovery and treat-
ment  system. The study area included the plant site and the area
between two streams in the vicinity of the plant site.
  Water was drawn from 25 of the 27 wells in the study area (Fig.
4). Wells located on the plant site were sampled Sept. 7.1983. Four
                                     "• LocltiOT ind runMr of motirtomq wftH
                           Figure 4
                  Location of Monitoring Sites
                                                                                           LEACHATE CONTROL      117

-------
 wells off the plant site (wells 24, 25, 26 and 27) were sampled on
 Jan. 6, 1984. These four wells had not been constructed when the
 September sampling was done. The water samples collected on
 Sept.  7 were analyzed for dissolved sulfates, total and dissolved
 lead and  total dissolved solids.  Specific  conductance, pH and
 temperature were measured in the field at the time of sampling. The
 water samples  collected on Jan.  6  were analyzed for total sulfates,
 total and dissolved lead, total and  dissolved chromium and specific
 conductance. Temperature and  pH  were measured in the field at
 the time of sampling. Sampling,  storage and preservation were per-
 formed  in accordance with procedures  recommended  by  the
 USEPA.'
   Streamflow  measurements were made and  water samples were
 collected at 10 locations along the streams (Fig. 4). These data were
 collected on Nov.  30. The water  samples were analyzed for total
 sulfates  and  dissolved  lead.  Specific  conductance,  pH  and
 temperature were  measured in  the  field. Sampling, storage and
 preservation were performed in accordance with procedures recom-
 mended by the USEPA.'
   Chemical analyses of the  water samples were performed using
 procedures recommended by the USEPA." The dissolved and total
 metals were  measured  by  preparing the samples according to
 general metals digestion procedures and then measuring  the consti-
 tuent metal content using atomic absorption spectroscopy. Sulfates
 were  measured  using  the turbidimetric  method. The dissolved
 solids (TDS) in  the water  samples  were  measured as the total
 filtered residue dried at 180°C. The results of the water analyses are
 given in Tables 5 and 6.
   Depth to water measurements for computing groundwater eleva-
 tions were taken  at each well at the time of sampling. The depth to
 water measurements were taken with a steel tape.  The water-level
 elevation in each well was determined by subtracting the measured
 depth to water from the reference point elevation on the top of the
                             well casing from which the measurement was made. Groundwater
                             elevations at the time of sampling are summarized in Table 5.
                               An earth electrical resistivity survey was performed using a Bison
                             2350B resistivity unit on, and in the vicinity of the plant site. In this
                             investigation, voltage drop was measured between two potential
                             electrodes placed in the earth resulting from  an applied current
                             through two other electrodes located outside, but in line with, the
                             potential electrodes. The standard Wenner electrode array was used
                             during  resistivity  profiling. Measurements  were made with elec-
                             trode spacings of 10, 30, 60 and 100 ft.
                               Three wells were constructed southwest of the plant site under
                             the supervision  of Engineering-Science personnel (Fig.  4).  These
                             wells were for use in hydrogeologic studies. The wells were con-
                             structed between Nov. 1 and Nov. 14. Wells numbered 24 and 25
                             were located in the vicinity of existing wells 1 A, 22 and 23. Well 25
                             was placed at approximately the same distance southwest of well 24
                             as wells 1 A, 22 and 23 are located northeast of well 24. In this man-
                             ner, well 24 could be used as the pumping well during an aquifer
                             pumping test and wells 1A, 22, 23 and 25 would be available for
                            observation  of drawdowns during testing. Well 26 was located ap-
                            proximately 1000 ft southwest of well 24 near the confluence of two
                            streams that are  hydraulically downgradient from the plant site.
                            This well was drilled to verify the existence of contamination iden-
                            tified in  that area by the geophysical survey. All three wells were
                            also for use in gathering lithologk data and sampling groundwater.
                              An additional well was constructed southwest of the plant site by
                            a local well driller under contract to the client.  This well, well 27,
                            was drilled during the  last  week in October. The well is approx-
                            imately 250  ft west of well 24 (Fig. 4).
                              A 24-hr aquifer pumping test of well 24 was performed on Nov.
                             19-20,  1983. Pumping began at 10:30 AM Eastern Standard Time
                            on Nov. 19 and  lasted  until 10:45  AM on Nov. 20. The pumping
                            rate was 4.17 gal/min. The pumped water was  discharged to the
                                                             TibleS
                                                 Summary of Groundwater Analyses
                  rHssolv..-.)    Total
                   Sulfrtt'i    Slllfat"    L"Ai\
                    (•q/ll    (mq/1)    (nq/l)
                                      Tot.ll   Dlnxulvod
                                              Uq/1)
                 Chrow v»)«
                  <«q/U
                              Sol l.Is
                              (••I/I)
        Con.1urt*r»T
          ('i«ho*>
              mi
          <*M. unit*)
                                  T-«,..
                                  (•(-)
                                                                          w.r-r
                                                                        Rl-virton
IA
3A
4A
5A
f,
7/\
H
'»
10
1 1
12
1 1
14
15A
1<1
17
IB
10
20
?l
•>,->.
31
S»ot.
Sept.
Sept.
Sept.
Sept.
Sept.
S«pt.
S.pt.
S'T>r. .
S^pc.
S/Tlt.
r.ept.
Snpt.
S«pt.
Slpt.
-,~pt.
r.npt .
Sept.
S"Pl.
Sfpt .
Sr.pt.
'i"nc .
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
7
                     810
                     600
                      52
                     750
                     140

                     440
                      24
                      23
                       1
                       1
                       2
                    1660
                     280
                       4
                     220
                     280
                     A40
                     5R9
0.14
 .97
 .If)
 .37
 .37


 .56
 .18
1.1
 .78
 .45


 .24
 .16
 .A3
1.4
 .22


 .12
 .2R
 .26


 .14


 .15
 . IA
0. 1?
 .25
 .04
 .15
 .IS


 .41
 . IA
 .00
 .11
 .20


 .18
 .21
 .51
 . V)
 . 14


 .00
 .24
 .10
 .21
 . 15
i , 11)0
  127
1.260
  76O
  110
  BO
   A;
   94
3. -17O
  555
   S7


   Rl


  542


  I 10


I , 170
I , 140
i.AOO
I, 42O
   89
1,2)0
  N64
   71
   15
   49
   11


   VI
   22
2.910
  MB
   4)


   54
  500
  M7


   17


I , I AH
i , I r>0
                        1.4
                         .0
                         . I
                         .7
                         .9

                         .1
                         .S
                         .4
                         .A
                         .1
                        4.2
                        4.4
                        4.1
                        a. 4
                        1.7
                        5.5
                                  17.n
I 7. /


17.2
"0.4
17.1
17.4
                                  17.1
                                  I 7.A
17. 1
17. i
                                           •vil.71
                                           •VI4. 15
 'VJO. 11
 •» i. in


IOJI.71
 O'lA.fll
 1'IA. 10
"1 .t.in. A
>'< -lin. A
A> 1 m . f,
' 1 1 'ii . A
1 , '110
1, 7OO
1,'',SO
.20
. 14
. IA
.V)
.11
. I?
.15
. 10
O.OO
<.05
<.05
. II
O.OO
. IIA
.OH
. IS
1, 400
1 , 120
il.H'iO
R, 100
1.7
l.fl
4. A
4. 1
1 7. 1
1 7. -1
1 7. 1
17. 1
118
         LEACHATE CONTROL

-------
                           Table 6
               Summary of Surface Water Analyses
    Measurement   Total   Dissolved    Specific   pH
Site   Date      Sulfate    Lead     Conductance (std.    Flow
 No.  (1983)      (mg/1)    (mg/1)        (unhos)   units)  (gpm)
1
2
3
4
5
6
7
8
9
10
Nov.
Nov.
Nov.
Nov.
Nov.
Nov.
Nov.
Nov.
Nov.
Nov.
30
30
30
30
30
30
30
30
30
30
368
332
251
50
15
559
598
415
293

0.
0.
0.
0.
0.
0.
0.
0.
0.

21
1 1
19
26
13
37
34
36
12

753
578
507
110
52
1000
1010
817
39

4.
4.
4.
4.
5.
4.
4.
4.
5.

3
2
4
8
3
2
8
4
1


41 .
27.
31.
12.
50.
30.
25.
6.
0.

6
1
2
0
4
0
0
5
0
plant wastewater pretreatment facility for treatment and discharge
to the city sewage lines. Wells 1A, 22, 23 and 25 were used as obser-
vation wells.
   Well construction details for the pumped well and the observa-
tion wells varied between wells. Wells 24 and 25 each have approx-
imately 34 ft of screen located in the upper one-half of the uncon-
solidated geologic deposits. Well 1A has 10 ft of screen located at
the water table of the unconsolidated deposits,  well 22 has a 5-ft
screened interval located in the center portion of these deposits and
well 23 is completed  as an open hole at the top of competent
bedrock.  A geologic cross-section showing the screen settings for
the wells used in the test is given in  Figure 5.
                 .-    r        '
                 ^MB4«J   wel used lor observing drawdowns. TD is
                 u_.	ffl   |0tai jjepin ol borng. WD is well deplh il
                     ID I   dilferenl than total depth.
                            Figure 5
          Well Arrangement Used for Aquifer Pumping Test

   Water levels were measured in the pumped well and each of the
observation wells. Water-level declines were observed in each of the
wells during pumping. The  maximum water-level  decline in the
pumped well was 12.7 ft. Maximum water-level declines in the
observation wells ranged between 0.18 and 0.53 ft.
   The water-level data for the pumped well and the observation
wells were used to compute the hydraulic properties of the uncon-
solidated deposits  in  the  vicinity  of the wells. The Theis  non-
equilibrium formula5 and water-level data for the observation wells
were used to compute transmissivity and storage coefficient. The
modified  nonequilibrium formula5  also was  used to  compute
transmissivity and storage coefficient.  Hydraulic conductivity for
the unconsolidated deposits was estimated based on the computed
transmissivity and estimated effective saturated thickness at each
well. The  results of the analyses are given in Table 7.
  Average hydraulic  properties for  the unconsolidated deposits
were estimated by averaging the results from the individual wells.
The average  hydraulic conductivity for these deposits is approx-
imately 49.5 gal/day/ft2. The average storage coefficient is about
0.0040. These numbers were computed by averaging the results for
wells 1 A, 22 and 23 to represent the hydraulic properties upgradient
from the pumped well, averaging the results for well 25 to represent
the hydraulic properties downgradient  from the pumped well and
using  these averages along with the computed hydraulic properties
at the pumped well to compute overall average hydraulic properties
for the test.
  The storage coefficient computed using data from well 1A ap-
peared to  be  anomalously high and was not used in the computa-
tions for the average value. This relatively high  storage coefficient
                                                                                   luttOQDQOOODODDG

                                                                                   X
                                                    \
                                          tiMxunoM
                                   • W.I.T l
-------
                                                             Table 7
                                                   Aquifer Pumping Test Summary
                Urtll
                Ho.
                        lqpd/ft>
                              Storage
                            Coefficient
Saturated
Thickness
  (feat)
 Hydraulic
Conductivity
 (gpd/ft )
 Method
  of
Analyai«•
                 Avq
                Av)
                          -. 990


                          5,600
                          1,410
                          3, 340
                          3,980
                          2,900
                          3,400
                          i, 380
                          4, 340
                          4, 400
                          4, 370
                               .0)6

                               .026
                               .0025
                               .002}
                               .0021
                               .0046
                                        .0029
                                                     86
                                                     66
                               .0053
                               .0050
                                        .0052
                                                     60
               34. i

               44.2
               39.7
               38.8
               46.3
               33.7
                                                                39. 5
                                                                36.}
               72.3
               73.3
                                                                72.8
                         Storage coefficient  not used  in
                         average
                         Storage coefficient  not used  in
                         a ve ra
-------
                                                                                    EXPLANATION

                                                                                 Off-atte property owned by
                                                                                 client and lot number

                                                                                 Preliminary well location
                            Figure 7
     Approximate Area of Offsite Groundwater Contamination
                            Figure 8
                   Preliminary Well Locations
ACKNOWLEDGEMENT

  This work was sponsored by the USEPA, Office of Waste Pro-
grams Enforcement.
REFERENCES

 1.  American Society for Testing Materials, "Standard test method for in-
    filtration rate of soils in field using double-ring infiltrometers (D3384)
    19:1982," Annual Book of ASTM Standards, Part 19, Natural Build-
    ing Stones, Soil and Rock, 1982.
 2.  American Society for Testing Materials, "Standard method for dry
    preparation of soil samples for particle-size analysis and determina-
    tion of  soil constants (D421)19:1982," Annual Book of ASTM
    Standards, Part 19, Natural Building Stones, Soil and Rock, 1982.
 3.  American Society for Testing  Materials, "Standard method for par-
    ticle-size analysis of soils (D422)19:1982,"  Annual Book of ASTM
    Standards, Part 19, Natural Building Stones, Soil and Rock, 1982.
 4. Fenn, D. et al., Use of the water balance method for predicting leach-
   ate generation from solid  waste disposal sites, USEPA, EPA/530/
   SW-168, 1975.
 5. Ferris, J.G., et at., "Theory of aquifer tests," U.S. Geological Survey
   Water Supply Paper 1536-E, 1962.
 6. Geraghty and Miller, Inc.,  Groundwater quality assessment at and in
   the vicinity of the Taylor Road landfill, Seffner, Florida, Consultants
   Report to the Division of  Public Utilities and Safety, Hillsborough
   County,  Florida, 1981.
 7. National Oceanic and Atmospheric Administration, "Climate of Plant
   City, Florida," National Oceanic and Atmospheric Administration
   Climatology of the United States, No. 20, 1976.
 8. Smith, W.N. and Talley, P.C., Operation and economic impacts of
   ground-water contamination, Taylor  Road  landfill,  a case study,
   County Commission and Department of Solid Waste,  1981.
 9. USEPA, Procedures manual for ground-water monitoring at  solid
   waste disposal facilities, SW-611,  1977.

10. USEPA, Methods for chemical analysis of water and waste  EPA-
   600/4-79-020,  1979.
                                                                                                   LEACHATE CONTROL
                                                                                                                                   121

-------
LABORATORY  INVESTIGATION OF PHYSICAL PROPERTIES
                 OF  FLEXIBLE MEMBRANE  LINERS (FML)
                                            BRIAN D. GISH
                                       WILLIAM E. WITHEROW
                                         Carlisle Syntec Systems
                                          Carlisle,  Pennsylvania
 TEST METHODS

 •Water Absorption ASTM D 471
    "Effects of Liquids on Rubber Property"
 •Dimensional Stability ASTM D 1204
    "Linear Dimensional  Changes  of Nonrigid  Thermoplastic
    Sheeting or Film at Elevated Temperatures"
 •Puncture Resistance FTMS 101B-2031
    U.S. Federal Test 101B Method 2031
 •Tensile Strength ASTM D 412 Die C
    "Rubber Properties in Tension"
 •Elongation ASTM D 412
    "Rubber Properties in Tension"
 •Brittleness Temperature ASTM D 746
    "Brittleness Temperature of Plastics Elastomers by Impact"
 •Tear Resistance ASTM D 624 Die C
    "Tear Resistance of Rubber"
 • Water Vapor Permeability ASTM E 96 Procedure BW
    "Water Vapor Transmission of Materials in Sheet Form"
 •Breaking Strength ASTM D 751 Grab Method
    "Tested Coated Fabrics"
 •Tearing Strength ASTM D 751 Tongue Tear Method
    "Testing Coated Fabrics"
                                                    80

                                                    70

                                                    60

                                                    50

                                                    40

                                                    30

                                                    ZO

                                                    10

                                                     0
                                                       -. INCREASE
                                                                     MEMBRASE TYPE
                                                                       Figure I
                                                              Nonreinforced Water Absorption
 TESTING

 Water Absorption
  Membrane specimens were immersed in distilled water according
 to ASTM D 471. Aging conditions were 7 days at 70 °C, and change
 in volume and change in mass after exposure were measured. The
 results of this testing are shown in Figures 1 and 2. All types of
 membranes increased in mass and volume with butyl rubber ex-
 hibiting  the least effect  and thermoplastic CPE exhibiting the
 highest increase in mass and volume.

 Dimensional Stability

  Membrane specimens were carefully measured for the length and
 width dimensions (machine and transverse directions, respectively).
 The specimens  were then placed in an  air-circulating oven  for 7
 days at 100°C.
  According to ASTM  D 1204,  after  the aging period, the
 specimens were  then remeasured and the percent changes were
 recorded. The results of this testing are shown  in Figure 3.  Butyl
 rubber exhibited the least average dimensional  change, and ther-
 moplastic CPE exhibited the greatest average dimensional change.
 All fabric reinforced membranes tested had less than 2% shrinkage.
                                                                             VOUJUt
                                                                             QUNCC
                                                      X INCREASE
                                                                     MEMBRANE TYPE
                                                                      Figure 2
                                                              Reinforced Water Absorption
 122
BARRIER TECHNOLOGY

-------
                    MACHINE      TRANSVERSE
                    DIRECTION       DIRECTION
                                                           AGED
                                                          28tl»70C
             AGED
            7d0116C
20 f
,.(
16
14
12

io|
      SHRINKAGE
                     EPDM          PVC "
                      MEMBRANE TYPE
                                                                           7. RETENTION OF ORIGINAL
                                                                 EPDM
                                                            MEMBRANE TYPE
                        Figure 3
            Nonreinforced Dimensional Stability
                                                             Figure 6
                                                    Reinforced Breaking Strength
                                                                                             AGED
                                                                                            28dO70C
                                                                        AGED
                                                                       7d9116C
800

700

600

500

400

300

200

 100

   0
     NEWTONS (N) FORCE
                       EPDM          PVC
                        MEMBRANE TYPE
                                                                             RETENTION OF ORIGINAL
                                                            TYL           VC
                                                            MEMBRANE TYPE
                          Figure 4
              Nonreinforced Puncture Resistance
                                                             Figure 7
                                                     Nonreinforced Elongation
                       AGED
                     2BdO70C
 AGED
7d0116C
 AGED
28H070C
 AGED
7d«M6C
160

140

120

100

  BO

  60

  40

  20

   0
     % RETENTION OF ORIGINAL
                      EPDU         BUTYL
                        MEMBRANE TYPE
                                                                            % RETENTION OF ORIGINAL
                                                                  CPE
                                                             MEMBRANE TYPE
                         Figure 5
               Nonreinforced Tensile Strength
                                                              Figure 8
                                                    Reinforced Elongation (Fabric)
                                                                                            BARRIER TECHNOLOGY
                                                                                                                                123

-------
                    ORIGINAL
                    UNAGED
                                   AGED
                                  71101180
           CELSIUS (C) TEMPERATURE
                            MEMBRANE TYPE
                                                                                               MEMBRANE TYPE
                            Figure 9
               Nonreinforced Brittleness Temperature
                                                                                      Figure 12
                                                                              Reinforced Tearing Strength
                                 »GfO
                               2SdO70C
                                   »CED
                                  7SOI16C
           -CELSIUS  TEMPERATURE
                                                                             r
                                                                005 ,
                                                               0045
                                                                004
                                                               0035
                                                                003
                                                               0025
                                                                002
                                                               0015
                                                                001
                                                               0005
                                                                  0
                                                                              METRIC PERM-CM
                          MEMBRANE TYPE
                                                                                               MEMBRANE TYPE
                            Figure 10
                 Reinforced Brittleness Temperature
                                                                                      Figure 13
                                                                        Nonreinforced Water Vapor Permeability
                            A^LL         ACCD
                          :eao?oc      rao\ 16C
             RETENTION OF ORIGINAL
                                                                       .0025
                                                                        .002
                                                                       0015
                                                                        .001
                                                                       0005
                                                                             METRIC PERM-CM
                            MEMBRANE TYPE
                                                                                              MEMBRANE TYPE
                           Figure 11
                  Nonreinforced Tear Resistance
                                                                                     Figure 14
                                                                         Reinforced Water Vapor Permeability
124
BARRIER TECHNOLOGY

-------
Puncture Resistance

  Membrane specimens were tested for puncture resistance accor-
ding to United States Federal Test 101B Method 2031. The results
of this testing are shown in Figure 4. All membranes showed good
puncture resistance with butyl rubber (1.562 mm.) exhibiting the
best puncture resistance. Puncture resistance which is dependent on
membrane thickness should be kept in mind when comparing these
results. Data for fabric reinforced membranes is incomplete.

Elongation

  Membrane specimens were tested for elongation retention after
aging conditions of 7 days at 116 °C and 28 days at 70 °C according
to SSTM D 412 or D 751. The results of this testing are shown in
Figures 7  and 8. The aging conditions of 7 days at 116°C were
generally more severe than those  of 28 days at 70 °C.

Tensile Strength/Breaking Strength

  Membrane specimens were  tested  for tensile/breaking strength
retention after aging conditions of 7  days at 116 °C and 28 days at
70 °C according to ASTM D 412 Die C or ASTM D 751. The results
of this testing are shown in Figures 5 and 6. All membranes showed
good retention of tensile/breaking strength after aging for both 7
days at 116 °C and 28 days at 70 °C except thermoplastic PVC which
exhibited the worst tensile strength retention after 28 days at 70 °C.
Brittleness Temperature
  Membrane specimens were tested for low temperature brittleness
according to ASTM D 746. Low temperature brittleness was also
determined after heat aging of 7 days at 116 °C and 28 days at 70 °C.
The results of this testing are shown in Figures 9 and 10. EPDM
and butyl rubbers and thermoplastic CPE and CSM exhibited good
retention of brittle point. However, thermoplastic PVC exhibited
the greatest effect of heat aging for brittleness temperature.

Tear Resistance/Tearing Strength
  Membrane specimens were  tested  for  tear  resistance/tearing
strength retention according to  ASTM D 624 Die C and ASTM D
751 tongue tear. Tear resistance  retention was also determined after
heat aging of 7 days at 116°C and 28 days at 70 °C. The results of
this testing are shown in Figures  11 and 12. Thermoplastic PVC and
CSM increased  in tear resistance after heat aging, and the other
membranes exhibited good retention of this property.

Water Vapor Permeability
  Membrane specimens were tested for water vapor permeability
according to ASTM E96 Procedure BW. The results of this testing
are shown in Figures 13 and 14. Butyl rubber exhibited the lowest
permeability (lowest amount of water vapor passing through an
area  per unit of  time)  while  thermoplastic PVC exhibited  the
highest permeability.
                                                                                         BARRIER TECHNOLOGY
                                                                                                                         125

-------
                                        CEMENT BARRIERS

                                              WAYNE S.  ADASKA
                                          Portland Cement Association
                                                   Skokie, Illinois
                                            NICHOLAS J. CAVALLI
                                              ICOS Corp. of America
                                               New York, New York
 INTRODUCTION

   Since the early 1970s several hundred slurry trench cutoff walls
 have been constructed to serve as underground barriers to the hor-
 izontal flow of water and other fluids. A major application has
 been in pollution control including the containment of hazardous
 wastes.
   One of the most commonly used slurry wall methods is the soil-
 bentonite (S-B) slurry trench. In this method, a bentonite-water
 slurry is introduced into the trench during excavation to provide
 side wall support. After the trench has been excavated to its re-
 quired depth, a mixture of soil, bentonite and water is placed into
 the  trench displacing the  bentonite-water slurry. The excavated
 soil  is  used in the backfilling operation; however, if it contains an
 excessive amount of contaminated or undesirable material such as
 cobbles  or  clay lumps,  a selected backfill  material may  be  re-
 quired.
   At sites where  wall strength is  important, waste compatibility
 with an S-B slurry mixture may be questionable or space limitation
 for mixing the S-B backfill adjacent to the trench may be a prob-
 lem. Alternative types of slurry walls are available. These include
 cement barriers such as  cement-bentonite slurry  trenches,  plastic
 concrete and structural diaphragm walls.
   In this paper, the authors describe the various types of cement
 barriers.  Information is presented on physical properties such as
 strength, deformability,  waste  compatibility and  permeability.
 The  authors also describe design and construction procedures  for
 four unique waste containment projects.
CEMENT-BENTONITE SLURRY TRENCHES

  For the cement-bentonite (C-B) slurry trench method,  cement
is added to the bentonite-water slurry just prior to its introduction
into the trench. In addition to serving as a stabilizing fluid to main-
tain an open trench during excavation, the C-B slurry forms the
permanent cutoff wall. The addition of cement generally increases
the cost of a C-B trench over a comparable S-B trench; however,
there are some distinct advantages with the C-B method:1
•The C-B method is not  dependent on availability or quality of
  soil for backfill.
•The C-B method is more suitable in trenching through weak soils
  where trench stability may be a concern. The C-B slurry begins to
  set within hours after excavation, thereby reducing the chance of
  failure.
                                                       •The C-B slurry sets up to a stiff claylike consistency. Trenches
                                                        may be cut through the wall without  sloughing. Construction
                                                        traffic may cross the trench after a few days.
                                                       •The construction sequence is more flexible. The C-B method per-
                                                        mits trench construction  in sections to meet site constraints. It
                                                        adapts to hilly surfaces where a step-type construction can be per-
                                                        formed. With  the S-B method, the long open trench necessary to
                                                        accommodate  the flat  slope of  the  backfill  normally requires
                                                        trenching continuously in one direction at a constant elevation.
                                                       •With a C-B slurry trench, construction may proceed during sub-
                                                        freezing temperatures. With the S-B method, special precautions
                                                        are required to keep the backfill from freezing.
                                                       •The width of a C-B trench is generally  less than for an S-B trench.
                                                       For the S-B method, the  trench must be wide enough to permit
                                                       free flow of the backfill material.
                                                       •With the C-B  method,  an area adjacent to the trench  is not re-
                                                        quired for mixing, thus making it more suitable in projects with
                                                        space limitations such as the crest of a dam. Also, cleanup is easier
                                                        with the C-B method.
                                                         Permeability  is one of  the most important  factors in  slurry
                                                       trench applications. Both laboratory and field tests indicate perme-
                                                       abilities of C-B slurry trenches approximately equal to 1 fl/yr
                                                       (10-• cm/sec).1
                                                         Since a C-B  slurry trench is not intended to support bending
                                                       moments or significant shear stresses, strength usually is not a pri-
                                                       mary consideration. The trench is designed to achieve a strength
                                                       equivalent  to that of the surrounding  soil. However, in projects
                                                       where slurry trenches are constructed  through unstable materials
                                                       such as peats and mine spoils, trench stability, especially during ex-
                                                       cavation, is a critical consideration. The cement-water ratio has a
                                                       significant effect on the strength of the C-B slurry trench. Also, as
                                                       with concrete,  strength  increases  with age. The effects of both
                                                       cement-water ratio and age on strength are shown in Figure la.
                                                         The deformability or compressibility of a slurry trench is impor-
                                                       tant when considering its application at waste disposal sites or in
                                                       seismic areas where displacements may occur. The slurry trench
                                                       must be able to accommodate the displacements without crack-
                                                       ing. A major  factor affecting the deformability of C-B  slurry
                                                       trenches is  strength. Laboratory tests indicate that higher strength
                                                       results in  a  stiffer less deformable wall. However,  even with
                                                       uniaxial compressive strength of SO psi, the C-B slurry exhibits a
                                                       high  strain capacity. The  relationship between ultimate uniaxial
                                                       compressive strength and triaxial strain at failure is shown in Figure
                                                       Ib.
126
BARRIER TECHNOLOGY

-------
        IOO
         80
      «  60
      i
      i
      S  40
         20
            o 7-day specimens
            • 26-day specimens
0.6


0.5
                O.IO    0.20    0.3O    0.40    0.50    0.60

                       Cement-water ratio C/W
         110
      x  120
      Ł


      I
                                  o Slurry sampled
                                   during construction
           0     5      10     15     20     25     30
                28-day trioxiol compressive strain at failure, €f %

                          Figure 1
   Typical Strength Deformation Test Results for Cement-Bentonite
                          Slurries


  For applications that involve contaminated water or exposure to
pollutants, it is important to determine the effect the waste has on
the slurry trench. For example, acids will dissolve the cement com-
ponents of  a C-B slurry trench. Sulfates  may  also be harmful;
however, the attack by sulfate soils or wastes may be reduced or
prevented by using cement containing a low tricalcium aluminate
(C3A) content. \n concrete design, Type II cement with a maximum
C3A content of 8% is used  for moderate sulfate exposure (150
to 1500 ppm), and Type V with a maximum CjA content of 5% is
used for severe sulfate exposure (1500 to  10,000 ppm).

Stroudsburg
  In October 1980, State of Pennsylvania officials observed black
tarry globules emanating from the  base  of a dike along Brodhead
Creek in Stroudsburg, Pennsylvania. The observed seepage was
in the approximate location of an old coal gasification plant which
generated coal tar as a by-product of the process.
  Prior to its closure in  1940, the  plant had disposed of coal tar
residue through an injection well located in the northwestern sec-
tion of the plant property. The well was constructed so that resid-
uals were injected into a gravel alluvium layer about 20 ft below
ground surface. Underlying this gravel  layer was  a layer  of fine
sand  which provided  an effective barrier to further downward
migration.
  In September 1981, the State presented its findings of an inves-
tigative study into the problem.3 The study indicated that the con-
tamination was confined generally to the gravel layer and had
spread over an area  approximately 8 acres. The report recom-
mended construction of a slurry trench cutoff wall to contain the
coal tar and prevent further migration into the streambed.  Also
recommended was the installation  of a recovery well system to
collect the coal tar wastes for removal.
  Because of the nature and extent of contamination, the  State
applied for and received funds for remedial work under the Super-
fund program. These funds were  appropriate on Nov. 9,  1981,
making Stroudsburg the first site to receive Emergency Superfund
monies.
  Under USEPA  supervision,  compatibility testing  was  con-
ducted to determine the most appropriate slurry wall composition.
The decision to use a C-B mixture was based upon three factors:
(1) the compatibility test results, (2) the lack of area for on-site mix-
ing of an S-B backfill and (3) the unavailability of local clays for
use in an S-B backfill.
  The slurry mixture, which has a design permeability of 10"'
cm/sec, consisted of 133 lb of bentonite and 345 Ib of cement per
cubic yard of slurry. The original trench design included the use of
a polyethylene  liner placed within the trench for added imperme-
ability. The length and weight of the material,  however, caused
problems during attempted installation and the material was  never
utilized.
  Construction of the slurry trench began on Nov. 25, 1981. The
completed trench is 648 ft long,  1 ft wide and 17 ft deep. It extends
down through the contaminated gravel stratum and is keyed 2 ft in-
to the  underlying sand layer.  The overall surface  elevation of the
trench is approximately 380 ft above sea level. The upstream end of
the trench is tied into  a sheet piling gate that is part of the existing
flood dike.  The downstream end  is tied  into  an impermeable
cement-bentonite  grout  curtain (Fig. 2).  The curtain was  con-
structed to form the final downstream segment because it was be-
lieved that trench excavation in proximity to the dike would have
imipaired  the dike's integrity. The grout curtain was installed by
pressure grouting through a series of vertical holes in the ground.
The curtain is approximately 50 ft long.4
  The trench was excavated with a backhoe along a narrow 11.5 ft
wide working platform.  During trenching operation, the contami-
nated material was separated and hauled by track-mounted bucket
loader to a small on-site storage basin. Periodically, the stored ma-
terial was loaded onto a sealed truck and transported to a disposal
facility in Niagara Falls, New York.
  The slurry trench was completed on Dec. 15, 1981, at which time
drilling began for the grout curtain. The cement-bentonite grout
curtain was completed within 7 days.
                                          Figure 2
                 Extent of Contamination and Slurry Wall Location at Stroudsburg Site
                                                                                           BARRIER TECHNOLOGY       127

-------
JFK Bulk Fuel Tank Farm

  Most slurry trenches constructed  for containment purposes re-
quire a clay or rock layer underlying the site to provide an imper-
vious stratum into which the trench can key. Petroleum products
such as fuel oil do not mix with and are lighter than water. As a
result,  they will float on top of the groundwater table. In this case,
it is not necessary to key the slurry trench into an impervious layer.
The bottom of the trench may simply extend a reasonable depth
below the minimum expected groundwater table.
  In 1980,  fuel oil was observed in an estuary channel adjacent to
a major fuel tank farm at Kennedy Airport in New York City. The
Port Authority  of  New  York and New  Jersey,  the  tank farm
owner, determined that spills within the tank farm over the years
had saturated the sandy soil and that oil was now draining into the
channel.
  The remedial action consisted of stabilizing the existing bank
with Till and  riprap and installing a slurry trench  to intercept  the
leaking oil (Fig. 3). Because a peat layer exists beneath the site, the
Port Authority specified the C-B slurry  trench method over  the
S-B method.  Since C-B slurry backfill has a lower unit weight than
S-B backfill,  less pressure would be exerted on the weak peat layer.
Restricted site conditions also favored the use of the C-B method.
Active high pressure fuel lines were buried as close as 3 ft to  the
trench, making an S-B backfill mixing operation over these pipes
very risky. At several locations along the 4000 ft long slurry trench,
piers  supporting utility and fuel lines  passed  over the trench.
Special care and equipment were required to excavate the trench in
these areas (Fig. 4).
                           Figure 3
     Typical Embankment Section After Remedial Work at Kennedy
                    Airport Fuel Tank Farm
                          Figure 4
    Working in Confined Area with Backhoe and 9-Ton Clamshell
   Laboratory tests were conducted to verify that the C-B slurry met
specified requirements. Permeability values for the C-B slurry mix-
ture having a cement-water ratio of 0.2 ranged  from 6.6 ft/yr at
7 days to 2.6 ft/yr cm/sec at 31 days. The permeability tests were
conducted  at an isotropic effective stress of 0.50 ton/ft2 using site
water as the effluent. Unconfmed compressive strengths ranged
from 0.35 to 0.8 ton/ft' for 7-day cured specimens.'
   As part of the remedial work, two 30,000 gal concrete retention
tanks collect contaminated  runoff and separate out the  oil pro-
ducts. The oil is recycled  to supply building heating fuel and fuel
for fire training exercises at the airport.

PLASTIC CONCRETE WALL
   In order  to meet  permit  requirements  for  two new coal-fired
units, Montana Power Co.  had to meet clean air standards for a
class one, pristine area. This was accomplished by Bechtcl Corpora-
tion whose engineers designed and tested  a new flue-gas scrubber
system for  the Colstrip Generating Station, Colstrip,  Montana.
The system uses a treated, hydrated dolomite lime to remove more
than 91%  of the sulfur dioxide.  While meeting  the Federal air
standard for emissions, the system produces a side stream of about
100,000 tons a year of scrubber waste.
   An effluent holding pond located approximately 4 miles east of
the plant was designed to accept the scrubber waste. The perimeter
of the pond was 16,536 ft long. To contain the waste, several pos-
sibilities were examined. A slurry cutoff wall was  eventually chosen
as the barrier method.
Contractor Option
   Prospective contractors were  prequalified and only those with
proven ability were allowed to submit proposals for the work. The
type of barrier and design was the contractor's option. One major
requirement was that the wall be 2.5 ft wide and  have a  maximum
permeability of I ft/yr.
  The geology of the site consisted of several near horizontal hor-
izons of tertiary  sedimentary rocks assigned to the Fort Union
Formation. This formation occurs as a series of interbedded, semi-
consolidated to consolidated carbonaceous sandstone, siltstone and
claystone with occasional coal beds.
  Since there was little to no soil to be excavated and no locally
available soils, the S-B method  utilizing soil as  backfill was con-
sidered the most expensive method. Also, because of the  subsur-
face conditions, a considerable length of time would be required to
excavate  the trench to the specified depth. With  the C-B method,
this time delay would cause  the C-B slurry mixture to set-up prior
to reaching final depth. Set retarding agents could be used, but be-
cause of  the variability in subsurface conditions  an exact set time
delay could not be established. Continually altering the mix  to facil-
itate construction would be neither desirable nor acceptable.
  The selected  contractor,  ICOS  Corporation of America, pro-
posed a plastic concrete wall 2.0 ft wide having  a permeability of
less than  1 ft/yr. This type of cutoff wall was determined to be the
easiest and  most economical. The method consisted of initially fill-
ing the excavation with a bentonite-water mixture to  keep  the side
walls  from  collapsing. The bentonite-water slurry was then re-
placed with a tremie concrete. The economy of the plastic con-
crete method became apparent in the design of the concrete back-
fill. Since the project was in  an area where fly ash is abundant and
inexpensive, the backfill was designed  using fly  ash  as a replace-
ment for  some of the  cement.
Testing

  Prior to construction, the concrete backfill was tested both for
permeability  and compatibility. The  testing procedure required
that backfill be permeated with water  to establish a base perme-
ability coefficient. Next, the specimen was to be permeated with
the scrubber waste composition (Table 1).  To determine the com-
patibility of concrete to the waste,  two pore volumes of effluent
were to be permeated and the results compared with the initial
water testing.
128
          BARRIER TECHNOLOGY

-------
                                                                                               Table 1
                                                                                  Typical Scrubber Waste Composition
                                                                   Component
                                                                   PH
                                                                   Conductivity
                                                                   Calcium
                                                                   Chloride
                                                                   Silicon Dioxide
                                                                   Magnesium
                                                                   Sodium
                                                                   Sulfate
                                                                   Total Dissolved Solids
                                                 Concentration*
                                                            8.3
                                                         30,000
                                                           500
                                                          1,000
                                                            30
                                                         10,000
                                                          4,798
                                                         49,494
                                                         69,805
                           Figure 5
      Permeability of Plastic Concrete Backfill Permeated with
                    Scrubber Waste Solution
  When permeated with water, the concrete specimen had the re-
quired permeability of  0.1 ft/hr.  When the same specimen was
permeated  with the scrubber waste liquid, the permeability de-
creased significantly to less than 10~" ft/yr (Fig. 5). This created a
new problem. With permeability this low and  getting lower with
time,  it was estimated it would take over 100  years to permeate
one pore volume.
                                                                  *A11 concentrations in mg/1 except pH (pH units) and conductivity Otmhos/cm).
  At that point, it was agreed the construction would begin and a
specimen of the backfill would be placed in a waste water bath as
a check for compatibility. After one year of being immersed,  an-
other permeability test would be run to determine if there was any
change in the permeability  of the  concrete.  At the time of this
paper, a year has not elapsed; therefore, these data are not avail-
able.
  Since the backfill chosen was concrete, a "panel" method of ex-
cavation was required. In this method, a series of alternating pan-
els referred to as primary panels are initially excavated. Following
completion of at least two adjacent primary panels, excavation of
an intermediate panel known as a secondary panel can begin. Due
to the variability in the  subsurface conditions,  a system of pre-
augering relief holes through the rock was conducted to facilitate
the clamshell excavation.

STRUCTURAL DIAPHRAGM WALL
  The majority of slurry cutoff walls for seepage control use either
the S-B, C-B or plastic concrete method.
  However, under certain  circumstances a  reinforced  concrete
slurry wall may be the most suitable method for waste containment.
This was the case on a recent classified site.
                                                           Figure 6
                                       Alignment of Structural Diaphragm Wall at Classified Site
                                                                                           BARRIER TECHNOLOGY
                                                          129

-------
                                                   2-0"
          S GREATER THAN 10' DEEP       *
                           Figure 7
         Typical Plan and Section of Structural Concrete Panel

   The site, confined by structures making containment both diffi-
 cult and expensive, is shown in Figure 6. Since the project was high-
 ly confidential at the bidding stages, neither the intended use nor
 final plan of the site was known by the invited contractors. All bids
 submitted for the project were rejected because they were over the
 owner's budget.
 Contractor Input

  After bids had been received and rejected by the owner, a meet-
 ing with  the owner was requested to discuss the total plan in order
 to lower the cost of the initial proposal. The feeling was that  if
 more parameters were known a change in design could be made to
 lower the cost of the project. During the meeting, it was revealed
 that the  owner intended the cutoff as a temporary measure; the
 final goal was to remove the waste to a secure landfill. The orig-
 inal design was to  install a C-B slurry wall barrier as the first of
 three separate contracts. The second phase was to excavate sheet
 piled pits in front of the wall replacing the contaminated soil with
 soil-cement in  10 ft wide alternating sections in front of the barrier
 to the top of the rock. This phase was not only very expensive, but
 also required a lot  of manpower to work in a hazardous environ-
 ment and handle hazardous material. The third contract would be
 to mass excavate the remaining soil and haul it to a secure landfill,
 replacing the soil with clean sand.
   Once the final plan was known, the C-B barrier was redesigned
 as a structural diaphragm wall capable of withstanding the earth
 pressure and surcharge from the adjacent building. The wall could
 now serve as an impervious barrier and retaining structure.
 Design

   The slurry wall was designed in the shape of a "T" (Fig. 7) to
 develop more friction against movement at the bottom and  reduce
 the size of the tiedown. The bracing system of the wall was de-
 signed using a tiedown that could be installed prior to excavation,
 thus minimizing exposure to the work force.
   The original geological information indicated a  sound, compe-
 tent rock formation; therefore, a minimal key was designed. How-
 ever, during the installation of the tiedowns, it was discovered that
 the rock had sand lenses throughout. This presented the possibility
 of waste channeling beneath the wall while the wall was serving as a
 barrier and allowing groundwater into the excavation during the
 removal of the waste. The implication for both cases would be eco-
 nomically disastrous, especially considering that the water entering
 into  the excavation would have to be hauled to a secure landfill
 for treatment. The discovery of these lenses led to  the design and
 installation of a cement grout curtain in the rock.
   During removal of the waste, no seepage was reported entering
 the excavated pit. The structural diaphragm wall, although more
 expensive  than the C-B wall, was less expensive overall since the
 second phase of the original design was eliminated.
 CONCLUSIONS

   The use of slurry walls has become an effective method for con-
 taining pollutants  at waste sites. The types of slurry walls include
 soil-bentonite and cement-bentonite slurry  trenches, plastic con-
 crete and structural diaphragm walk. Some  sites have special con-
 straints or requirements which may influence the type  of  slurry
 wall used.
   For  complex or  unique projects, the slurry wall contractor may
 be best qualified to determine the type of slurry wall and construc-
 tion method. In these situations, a performance rather than pre-
 scription type of specification is preferred.  This allows an exper-
 ienced contractor to provide the most appropriate and economical
 solution to the problem.
REFERENCES

1. Ryan,  C.R., "Slurry Cutoff Walls, Design and Construction," Re-
   source Management Products Slurry Wall Technical Course. Chicago,
   111..Apr. 1976.
2. Lolgani, K.L. and Kleiner, D.E., "Cement-Bemonite Slurry Trench
   Cutoff Walls,"  Proc, Seventh Pan-American  Conference on SoU
   Mechanics and Foundation Engineering,  June 1983.
3. Pennsylvania Department of Environmental Resources, Brodhead Ex-
   tent of Contamination Report, Sept. 1981,  Department of Environ-
   mental Resources, Wilkes Barre, PA.
4. Cochran. R.S. and Yang, E., Case Studies 1-23: Remedial Response
   at Hazardous  Waste Sites, EPA-540/2-84-0026,  U.S. Environmental
   Protection Agency, Washington, D.C., 1984.
5. Ladd,  R.S., "Laboratory Testing  Program, Report No. 1," Wood-
   ward-Clyde Consultants, Clifton, NJ, May 1980.
130
         BARRIER TECHNOLOGY

-------
BAKKIER-LEACHATE COMPATIBILITY:  PERMEABILITY OF
     CEMENT/ASPHALT EMULSIONS AND CONTAMINANT
                RESISTANT BENTONITE/SOIL MIXTURES
                               TO ORGANIC  SOLVENTS
                                         DAVID C. ANDERSON
                                              ALICIA GILL
                                          WAYNE CRAWLEY
                                    K.W. Brown and Associates, Inc.
                                          College Station, Texas
INTRODUCTION

  Organic solvents have been shown to increase the permeability of
flow barriers constructed with compacted  clay.'-11 In addition,
acids, bases, salts and organic liquids  have been reported to in-
crease the permeability of soil/sodium bentonite mixtures.12-13 Nev-
ertheless, barrier materials are widely prescribed to contain concen-
trated leachates, and millions of dollars are being spent to construct
sodium bentonite cut-off walls around Superfund sites.
  Some consultants and barrier material suppliers have suggested
that "real world" leachates contain only low "parts per million" of
organic chemicals. To support these claims, it has been stated that
the concentration of most organic chemicals in leachate is limited
by their low solubility in water. In fact,  many of the most common
and most toxic organic solvents are very sparingly soluble in water,
but this does not mean that those solvents only exist at very low
concentrations. Rather, it means that disposed organic solvents will
often be present as a separate, immiscible and concentrated organic
liquid layer.
  Although large pools  of organic liquids are rare at newly con-
structed disposal facilities, they are  common at Superfund sites.
Some have suggested there will almost always be enough water pre-
sent in disposal sites to solubilize even  the most sparingly soluble
organic liquids. However, to better understand the large volume of
water required to solubilize immiscible organic liquids, one should
consider the volume of  seawater required to solubilize oil slicks.
For those organics having a solubility limit of parts per million, it
would obviously require millions of gallons  of water to solubilize
only a few gallons. Although miscible  organic liquids are usually
diluted by water within a disposal site,  even  these miscible liquids
may be found in localized high concentrations if little mixing has
taken place.
  The USEPA banned disposal of "free" liquids in landfills in
1982.14 Immiscible organic liquids may, however, be present in sites
where no "free" organic liquids have been disposed. Leaching
studies were conducted using a mixture  of sand and sorbents com-
bined with 5% xylene, by weight." This percentage of xylene is far
below the concentration at which "free" xylene (as defined by the
USEPA) would be present. After passage of two pore volumes of
water through the  mixture, as much as 46% of the xylene was
released. This xylene was present as a floating immiscible liquid
layer on the water that drained from the sand-sorbent mixture.
  Considering the previous discussion, there seems to be a real
need to set the following goals concerning remedial actions at aban-
doned sites:
•Increase the use of waste  destruction, detoxification and treat-
 ment technologies  in remedial actions at Superfund sites. This
 should coincide with a reduction in  the use  of containment tech-
 nologies at these sites.
•For Superfund sites where it is not technically feasible to  destroy
 or detoxify the waste, consider the following:
 -compatibility tests between the most concentrated leachate at the
 site and the barrier material proposed to contain the leachate
 -examination of alternatives to sodium bentonite as the contain-
 ment material where acids, bases, salts or concentrated organic
 liquids are present in the leachate
  Some alternative materials that have been suggested for contain-
ment walls include the following:
•Extruded clay minerals
•Non-bentonite clay minerals (e.g., kaolinite, attapulgite, etc.)
•Calcium bentonite
•Silica based gels and formulations
•Vertical membranes (HDPE, PVC, etc.)
•Contaminant resistant bentonite
•Cement/asphalt emulsions
  In this paper, the authors report the findings of compatibility
tests conducted on two of the above materials: a contaminant-
resistant bentonite/soil mixture and a cement/asphalt emulsion.

PREVIOUS STUDIES
  Contaminant resistant bentonites have been produced by several
bentonite  mining  companies, drilling mud  producers  and even
chemical companies in  the United States and England.12-13'16'17
While some have  suggested that there is a contaminant resistant
bentonite for every containment problem, few independent studies
have been conducted to verify these claims.
  In a study evaluating the permeability and chemical resistance of
cement/asphalt emulsion fly ash mixtures, permeabilities decreased
with time  when the mixtures were permeated by water, an acidic
solution and a neutral salt solution.18 The mixtures consisted of
36% cementitious solids (fly ash and portland cement  in several
proportions),  52% asphalt emulsion and 12% water, by weight.
The permeabilities of the mixtures initially averaged 6 x 10 ~8 cm
sec-1.
MATERIALS AND METHODS

  Compatibility  test  results  are normally  reported  as the
permeability of a material during passage of two or more  pore
volumes of the permeant liquid. One pore volume is equal to the
total volume of pores within a sample. If, for instance, a 1000 cm3
sample  had a porosity of 0.4, then one pore volume would equal
400 cm3. Porosity of a sample can usually be calculated as follows:
              .= Porosity
                                                  (1)
where:      PD
    BD = density of the bulk sample on an oven dry weight basis
         (gm/cm3)
    PD =  density of solids in the bulk sample on an oven dry
         weight basis (gm/cm3)
                                                                              BARRIER TECHNOLOGY
                                                                                                          131

-------
With a bulk density of 1.59 gm/cm3 and a particle density of 2.65
gm/cm3, the porosity of the sample would be 0.4.
  Determining the pore volume of clay  is  relatively straightfor-
ward; however, this is not the case with materials such as asphalt
emulsions. A significant portion of the asphaltic material would
vaporize in any attempt to obtain an oven dry weight. Consequent-
ly, data are presented in this report in terms of permeability plotted
over cumulative time instead of on a pore volume basis. This allows
comparison between  results  obtained with the cement/asphalt
emulsion and the contaminant resistant bentonite/soil mixtures.

Double-Ring Permeameters Used in the Studies
  Permeability tests have been conducted on barrier materials us-
ing either fixed-wall or flexible-wall permeameters.' In fixed-wall
permeameters, sidewall leakage can be substantial." In flexible-
wall permeameters, confining  pressure may reduce the permeability
of soft or remoldable barrier materials.1
  Double-ring permeameters reduce the problems of sidewall flow
and confining pressure. Confining pressures are not used with these
devices, and flow near the sidewall is separated from flow through
the central portion of the sample. Double-ring permeameters were
originally  suggested by  McNeal and Reeve as  a method of
eliminating boundary flow errors." More recently, it has been sug-
gested that divided outflow  permeameters (such  as double-ring
devices) may give more reliable permeability results  than the tradi-
tional fixed-wall permeameter.11
  Double-ring permeameters  were used in all permeability  tests
discussed in this report. The permeameters were  15 cm in diameter
and had double-ring base plates (Figs. 1 and 2).  These base plates
were  fitted with drainage  layers consisting  of  either a sheet of
geotextile (cement/asphalt emulsion) or both geotextile and  sand
layers (contaminant resistant bentonite/soil). All permeability tests
were  conducted at  hydraulic  gradients of 36 except for approx-
imately the last 100 days of the cement/asphalt emulsion tests.
These tests were completed at an  hydraulic gradient of 72. All
permeability values  given in this paper were obtained from the in-
ner chamber of  the double-ring permeameters. In this report, the
terms standard leachate and water are used interchangeably.
Contaminant Resistant Bentonite/Soil Mixtures

  A slurry composed  of contaminant  resistant bentonite and dis-
tilled  water was  prepared as suggested by the supplier using 90 kg
bentonite/m3 of water." Distilled water was mixed in a blender (on
low speed), and  the clay was  slowly added until a 9% solution of
clay was obtained. The slurry was poured first through a screen to
ensure removal of any large aggregates of clay and then into the
                         PRESSURE INLET
                                            TEFLON
                                             GASKETS
                                            INNER
                                            RING
                                                ER
                                            OUTLET
                                                   OUTLET FOR
                                                    INNER RING
                                         OUTLET FOR-
                                         OUTER RING
        INNER OUTLET
DRAINAGE LAYERS
                          Figure 1
  Schematic of Double-Ring Permeameter for Testing Barrier Material
                                                                 Figure 2
                                            Details of the Base Plate to the Double-Ring Permeameter
15-cm double-ring permeameters. The filter cake that forms on the
sidewall of a slurry trench was simulated by allowing the clay to set-
tle out and form a 1-cm thick layer over an underlying sand layer."
  Calcareous smectitic clay soil was added to the nine percent ben-
tonite slurry until a homogeneous paste was obtained. The soil-
slurry paste had an average slump of  12.5-15 cm (as measured with
a standard concrete slump cone). Moisture content  and slump of
the soil-slurry  mixtures averaged 65%  (dry weight) and  14 cm,
respectively.  The  soil-slurry  mixture  was  then  poured  into
permeameters displacing the nine percent bentonite slurry."
  Permeability  tests began by permeating the samples with a stan-
dard leachate (distilled water  containing 0.005-0.010 N CaSO^.
After one month of permeation by water, an additional 5-cm thick
layer of soil-slurry mixture was added to each permeameter. All
soil-slurry mixtures were then permeated  with the standard leachate
for one additional month to obtain stable baseline permeability
values.11 At that time, the standard leachate was replaced with
either a non-polar organic solvent (xylene) or a polar organic sol-
vent (methanol). The physical and chemical properties of the per-
meant liquids used are reported in Table 1.
Cement/Asphalt Emulsion

  Ingredients of  the cement/asphalt emulsion were  thoroughly
stirred in the following proportions as suggested by the supplier"
1.  36% asphalt emulsion
2.  52% clean sand
3.  8% clean water
4.  4% type  1 portland cement
The mixture was poured into the permeameter molds immediately
after mixing to assure that the sand and cement remained in suspen-
sion. In a previous study, the material was found to shrink  slightly
as it cured into  a plastic solid." To obtain a good seal between the
material and permeameter sidewall,  it was necessary to cure the
material under an overburden pressure of 0.5 lb/in.2 After the first
24 hr of curing,  the overburden pressure was removed and all
permeability tests were conducted without overburden pressure.
  All cement/asphalt emulsion mixtures were initially permeated
with standard leachate (distilled water containing 0.005-0.010 N
CaSOJ for at least one month or until stable baseline permeability
values were  obtained.  At that time, the standard  leachate was
replaced with either a polar organic liquid (methanol) or one of two
non-polar organic liquids (xylene or creosote oil). (Table 1.)

RESULTS

  Permeabilities of both  the contaminant resistant bentonite/soil
mixture and  the  cement/asphalt emulsion were evaluated using
water, xylene and methanol as the permeant liquids. In addition,
the  cement/asphalt emulsion  was permeated  with  creosote oil.
Relatively low  permeabilities  were obtained with both  materials
when they were permeated by water.  Results were, however, strik-
ingly different  when the barrier materials were permeated by
organic  solvents.
 132
          BARRIER TECHNOLOGY

-------
                            Table 1
      Physical and Chemical Properties in the Liquids Studied"
   Property
   Water      Xylene    Methanol
(Standard   (Nonpolar    (Polar
 Leachate)   Solvent)    Solvent)
Density e.20°C
(gm/cnf3)
Viscosity e 20eC
(centl poise)
Dielectric Constant
0 20°C
Dipole Moment
( debyes )
Water Solubility
e 20°C (g L'1)
Molecular Weight
Melting Point (°C)
Boiling Point (°C)
0.98
1.00
80.4
1.83
—
18
0
100
0.87
0.81
2.4
0.40
0.20
106
-47
139
0.79
0.54
31.2
1.66
nlsclble
32
-98
6k
NOTE: Creosote oil, which was also studied, is a liquid distillation product (200-400 °C) of coal tar
consisting mainly of non-polar aromatic hydrocarbons along with some acidic and basic compounds.
It is immiscible, has a density and viscosity greater than water and a dielectric constant much less
than water.
Contaminant Resistant Bentonite/Soil Mixtures
  Permeability values for the contaminant resistant clay based
slurry ranged from 3 to 8 x 10 -8 cm/sec when permeated by water
(Figs. 3 and 4). When this material was permeated by either an im-
miscible or  miscible organic liquid, however,  it underwent large
permeability increases.
  Samples permeated by methanol exhibited stable permeabilities
during the first 20 days of exposure to the polar organic liquid (Fig.
3). By the end of the first month, however, permeability of the clay-
based slurry had increased from about 5  x 10 ~8 cm/sec to greater
than 1  x 10-5 cm/sec.
  Clay-based slurry samples began to undergo permeability in-
creases within the first few days of permeation by xylene (Fig. 4).
One of  these  samples underwent a  more  than two orders of
magnitude increase in permeability within the first week of permea-
tion by xylene. The other sample of clay-based slurry completely
failed  overnight. This  failure and  subsequent desiccation  from
direct exposure to pressurized air made it impossible  to calculate
the final permeability of the sample.
  Cement/Asphalt Emulsions

  Permeability values for the cement/asphalt emulsion (during the
first month of permeation by water) ranged  from  4  x  10 ~9
cm/sec  to less than 1 x 10~10 cm/sec (Figs. 3-5). As seen in an
earlier study, the permeability of this material decreased  over the
entire 30 days during which it  was permeated by water."
  After stable baseline permeabilities were obtained, samples of the
cement/asphalt  emulsion  were permeated by either  polar
(methanol) or non-polar (xylene and creosote oil) organic liquids
(Figs. 3-5). Permeability values thereafter varied somewhat but, in
all cases, the permeability remained below 5 x 10~9 cm/sec.
  Samples permeated by methanol had permeabilities that varied
over time from slightly less than 5 x 10 ~10 cm/sec to  slightly
greater than 3  x 10- u cm/sec (Fig. 3). After 330 days  of exposure
to methanol, permeability of the samples appeared to stabilize in
the range of 1 to 2 x 10-10 cm/sec.
  Samples permeated by xylene had permeabilities that varied over
time from 2x IQ-'cm/sec to 4 x 10-" cm/sec (Fig. 4). After 280
days of exposure to xylene, the range in  sample permeability de-
creased to between 1 and 6 x  10 ~10 cm/sec.
  Permeabilities varied from Ito3 x 10 ~9 cm/sec during the first
100 days of exposure to creosote oil. The range of permeability ex-
hibited by these samples increased to between 5 x  10~9 cm/sec
and 5 x  10 -10 cm/sec during the second 100 days of exposure to
creosote  oil (Fig. 5).

CONCLUSIONS
  The two alternative cut-off wall materials evaluated in this study
were  a  contaminant  resistant  bentonite/soil  slurry   and  a
cement/asphalt emulsion. Permeability values of these materials to
water were 5 x 10 ~8 cm/sec for the slurry and less than 5 x 10-9
cm/sec  for  the asphalt emulsion.  In  other  words,  the
cement/asphalt emulsion was  at most one tenth as permeable to
water as  was the contaminant resistant bentonite/soil based slurry.
  When permeated  by organic  liquids,  differences   in  the
permeabilities of the  materials greatly increased. Contaminant
resistant bentonite/soil mixtures underwent permeability increases
of from two to three orders of magnitude in three out of four cases.
Permeability increases in the clay based slurry occurred within
    no"7-
 a
-E
         WATER  '  METHANOL
                                      CEMENT-
                                      ASPHALT-
                                      EMULSION
                         SOIL-CONTAMINANT
                         RESISTANT
                         BENTONITE
      60 40  20  0  20 40  60 BO 100 120 140 160 180 ZOO 220 240 260 290 3OO 320 MO
                           TIME (DAYS)

                           Figure 3
   Permeability of Cement-Asphalt Emulsion and Soil-Contaminant
            Resistant Bentonite to Water and Methanol
                                                                    \  i.ID-*--
                                                                             WATER    XYLENE
                                      CEMENT.
                                      ASPHALT
                                      EMULSION
SOIL-CONTAMINANT
RESISTANT
BENTONITE
                                                                                                      REP I
                                                                                                      REP 2
                                                   60 »0 20  0  20 40 60 80 IOO 120 140 ISO ISO 20O 220 240 260 280 300 320^40
                                                                        TIME (DAYS)

                                                                       Figure 4
                                               Permeability of Cement-Asphalt Emulsion and Soil-Contaminant
                                                         Resistant Bentonite to Water and Xylene
                                                                                             BARRIER TECHNOLOGY      133

-------
                                       CEMENT-
                                       ASPHALT
                                       EMULSION
                            Figure 5
  Permeability of Cement-Asphalt Emulsion to Water and Creosote Oil
 either one week (for the xylene treated samples) or one month (for
 the  methanol treated samples).  When cement/asphalt emulsions
 were tested using permeameters, liquids and hydraulic gradients
 similar to those used with the clay based slurries, the  following
 results were obtained:
 •After permeation by methanol for 11 months, the  permeability
  of cement/asphalt emulsion samples had decreased slightly from
  permeability values obtained with water.
 •After permeation by xylene for over 9 months, the  permeability
  of cement/asphalt emulsion samples had also  decreased slightly
  from values obtained with water.
 •After permeation by creosite oil for 5 and 7 months,  permeability
  values of cement/asphalt emulsion samples were more variable
  than, but approximately the same as, for water.
   Several  studies have found that the permeabilities of clay liners
 and sodium bentonite slurry walls may be increased as the result of
 exposure  to concentrated leachates. Since Super fund sites often
 contain  such  leachates,  the following  approaches  should  be
 adopted for remedial action at these sites:
 •Reduce the role of contaminant technologies in remedial actions
  at Superfund sites
 •Increase the  role of waste destruction,  detoxification and treat-
  ment technologies in remedial actions at Superfund sites
 •For Superfund  sites where the  approaches suggested in 1 and 2
  above are not technically feasible, alternatives to sodium benton-
  ite  slurry walls should be examined
   In summary, the following  points are important:
 •Cement/asphalt emulsion exhibited a  permeability  to water an
  order of magnitude lower than that exhibited by the contaminant
  resistant  bentonite clay based slurry.
 •The clay based slurry exhibited  large permeability increases when
  permeated by organic solvents. These increases occurred within
  one week for the non-polar solvent (xylene) and  within four weeks
  for the polar solvent (methanol).
•When tested  using identical permeameters,  similar  hydraulic
 gradients and the same organic  solvents as were used on the clay
 slurry, cement/asphalt  emulsions  exhibited  either  steady  or
 slightly decreased permeabilities after greater than nine months
 of  testing. In addition, cement/asphalt  permeated  by  creosote
 oil  for seven months exhibited permeabilities of less than 5 x
  10-9 cm/sec.

REFERENCES
  1. Anderson, D.C. and Jones,  S.O., "Fate of Organic Liquids Spilled
   on Soil." Proc. of the National Conference on Hazardous Waste and
   Environmental Emergencies,  Houston. TX, March, 1984, 384-388.
 2. Brown, K.W., Thomas, J.C. and Green, J., "Permeability of Com-
    pacted Soils to Solvent Mixtures and Petroleum Products."Proc. of
    the Tenth Annual Research Symposium. Land Disposal of Hazardous
    Waste. EPA 600/9-84-007. 1984, 124-137.
 3. Daniels,  D.E. and Anderson,  D.C.. "Fixed-Wall vs.  Flexible-Wall
    Permeameters. Proc.  of the Symposium on Impermeable Barriers for
    Soil and Rock.  American Society for Testing and Material*. Denver,
    CO, June, 1984.
 4. Brown, K.W. and Anderson,  D.C.. Effects of Organic Solvents on
    the Permeability of Clay Soils. USEPA, Washington, DC, EPA-600/
    2-83-016, 1983.
 5. Anderson, D.C.  and Jones,  S.G.,  "Clay  Barrier-Leachaic Inter-
    action,"  Proc.  of the National Conference on Management of Un-
    controlled Hazardous Waste Sites. Washington, DC, Oct., 1983.
 6. Brown, K.W., Green, J. and Thomas. J.  "The Influence of Selected
    Organic  Liquids on the Permeability of Clay Liners." D.W.  ShulLz
    (ed.),  Land Disposal,  Incineration and Treatment of Hazardous
    Waste. Proc. of the Ninth Annual Research Symposium, EPA-600/
    9-83-002, 1983.
 7. Anderson, D.C.,  "Does Landfill Leachate  Make Clay Liners More
    Permeable?". Civil Eng., 32(9). 1982, 66-69.
 8. Anderson, D.C.,  Brown, K.W.  and Green, J.,  "Effect of Organic
    Fluids on the Permeability of  Soil Liners," D.W. Shultz (ed.) Land
    Disposal of Hazardous Waste.  EPA-600/9-82-002, 1982, 179-190.
 9. Anderson, D.C. and Brown, K.W., "Organic Leachate Effects on the
    Permeability  of  Clay  Liners," D.W.  Shultz (ed.). Land Disposal:
    Hazardous Waste. EPA-600/9-81-002b. 1981, 119-130.
10. Anderson, D.C.,  Organic Leachate Effects  on the Permeability of
    Clay Soils. M.S.  Thesis, Texas A&M University, College Station,
    TX. 1981.
11. White, R., Remolded Soil Samples for Proposed Waste Landfill Site
    North of Three Rivers, Texas.  Trinly  Engineering Testing Corpora-
    tion, Report No. 76791, Corpus Christi. TX, 1976.
12. D'Appolonia, D.J., "Soil-Benlonite Slurry Trench Cutoffs. J.  of the
    Geotechnical Eng.. ASCF. 106, *GT4. 1980, 339-417.
13. Hughes,  J., "A Method for  the Evaluation of  Bentonites as Soil
    Sealants  for the Control of Highly Contaminated Industrial Wastes."
    Proc., 32nd  Industrial  Waste Conference  at Purdue University,
    Lafayette, IN. Ann Arbor Science Publishers, Inc., Ann Arbor, Ml,
    1977, 814-879.
14. USEPA.  "Standards for Owners and Operators of Hazardous  Waste
    Treatment, Storage, and  Disposal Facilities," Federal Register 47,
    No.  143, July 26,  1982, 32274-32388.
15. Anderson, D.C. and Brown, K.W., Quantification of the Release of
    Xylene From Solids: Implications for Use of Absorbents to Remove
    Free Liquids From Landfills. Unpublished Research, 1984.
16. Miles, M.M. and Boyes, R.G.H., "Slurry Trenching Developments,"
    Civil Eng.. London. England. Apr., 1982, 51-52.
17. Garner, K., "Contaminant Resistant Bentonite." Civil Eng., London,
    England, Apr.,  1982.
18. Diamond, S., Cement Asphalt Emulsion—Flyash Mixes as Slurry Wall
    Components.  Report prepared for Slurry Systems, Inc., East Chicago,
    IN,  1982.

19. Anderson, D.C.,  Effects of Organic  Solvents on Clay Liner-Con-
    taminant  Resistant Bentonite Slurry Mixtures.  Report  for USEPA,
    Cincinnati, OH, 1983.

20. McNeal,  B.L. and  Reeve,  R.C.,  "Elimination of Boundary-Flow
    Errors in Laboratory Hydraulic Conductivity Measurements," Soil
    Sci.  Soc.  Am. Proc.. 28. 1964. 713-714.
21. Mclntyre,  D.S., Cunningham, R.B., Vatanakul,  V. and  Stewart,
    G.A., "Measuring Hydraulic  Conductivity  in Clay Soils: Methods,
    Techniques, and Errors," Soil Science. 128,  1979, 171-183.
22. Brinkman, R.,  Personal Communication. Technical Representative
    for American Colloid Company, Evanston, IL, 1982.
23. Anderson, D.C., Crawley, W.  and Zabcik, D.,  "Effect of Various
    Liquids on Clay Soil-Bentonite Slurry  Mixtures," Proc. of the Sym-
    posium on Impermeable Barriers for Soil and Rock, American Society
    for Testing and Materials, Denver, CO, June, 1984.
24. Zlamal,  F. Personal Communication. Manager of Technical Services
    for Slurry Systems, Inc., Gary, IN, 1983.
 134
           BARRIER TECHNOLOGY

-------
A LABORATORY  TECHNIQUE FOR ASSESSING  THE IN  SITU
             CONSTRUCTABILITY OF A BOTTOM BARRIER
                                  FOR  WASTE ISOLATION
                                      THOMAS P. BRUNSING, Ph.D.
                                           Canonic Engineers, Inc.
                                              Chesterton, Indiana
                                           RAY B. HENDERSON
                                              Foster-Miller, Inc.
                                           Waltham, Massachusetts
INTRODUCTION

  Block displacement, an innovative isolation technique developed
during the last few years, has generated the need for scaled testing
of the process. Under  an SBIR grant from the National Science
Foundation,  a facility to conduct this model testing has been
developed at Foster-Miller, Incorporated (FMI)  in Waltham,
Massachusetts. The  facility permits one-tenth scale modeling of
specific chemical waste site geologies to evaluate Block Displace-
ment application under those specific site conditions.
Process Description

  Block displacement is a method for vertically lifting a large mass
of earth. The technique produces a fixed underground physical bar-
rier placed around and beneath the earth mass.  The barrier is
formed by pumping slurry, usually a soil, bentonite and water mix-
ture, into a series of notched injection holes. The resulting barrier
completely isolates the earth mass from  groundwater migration
(Fig. 1). The barrier material should be compatible with in situ soil,
groundwater  and leachate conditions.
  A bottom barrier is formed when lenticular separations extend-
ing from horizontal notches at the base of injection holes coalesce
into a larger separation beneath the ground being isolated (Fig. 2).
Continued pumping of slurry under pressure produces a large uplift
force  against the bottom of the  block  and  results in vertical
displacement proportional to the volume of slurry pumped.
                          SLURRY
                          INJECTION
                           I   I
    GROUNDHATER  ^f.
    LEVEL
     POSITIVE SEAL THROUGH
     INJECTED BENTONITE
     MIXTURE
PERIMETER
^SEPARATION
1 (1 4
INJECTION
DHOLES X
UPLIFT
X PRESSURE \
MM HH

IMUl/
                        FRACTURED BEDROCK
                        Figure 2
           Creating Separation to Induce Displacement

  A perimeter barrier is constructed in conjunction with the bot-
tom barrier either prior to, during or following bottom barrier con-
struction. The perimeter barrier can  be constructed by various
means including slurry wall, vibrating beam or jet grouting tech-
niques. If constructed prior to bottom separation, the perimeter
wall can be used to ensure a favorable horizontal stress field for
proper orientation of the  propagating bottom separation. A
perimeter separation would first  be  constructed and then sur-
charged to increase the horizontal stress field in the formation (Fig.
2). Surcharge is additional pressure transmitted to the fluid slurry
in the perimeter separation either by raising the  slurry field level
above ground level or by placing a seal in the perimeter separation
and pressurizing the slurry below the seal.
  The Block  Displacement Method can be used to increase the
width of an initially thin perimeter barrier such as might be con-
structed by vibrating beam or jet grouting techniques. To increase
perimeter width by means of block lift, the perimeter must be con-
structed on a slight angle off vertical  prior to the lift. The  slight
angle, *, off  vertical tapers inward toward the block center. Up-
ward displacement, d, of the block resulting from injection  along
the bottom barrier will then increase the perimeter separation, W0,
to the desired barrier thickness, W, according to:
                                                                W = d sin $ + W0
                                                                                                              (D
                        Figure 1
              Block Displacement Barrier in Place
Field Experience

  A field demonstration of this process has been completed, under
USEPA sponsorship, adjacent  to a  National  Priority chemical
waste site near Jacksonville, Florida. The demonstration "block"
                                                                                BARRIER TECHNOLOGY
                                                                                                              135

-------
                          -BOTTOM SLUR" BARHIER
                             StCTIOH A-n


                            Figure 3
                     Final Block Displacement
was 60 ft in diameter by 23 ft deep and was composed of uncon-
solidated marine sediments.1
  The test program successfully demonstrated the fundamental
aspects of bottom barrier construction. A perpendicular cross sec-
tion of the displacement  of the earth mass is shown in Figure 3.
Core samples were taken to verify the continuity of the bottom bar-
rier in situ. Barrier thicknesses ranging from 5 in. to  12 in. were ob-
tained.
  The testing described in this  paper was intended,  in  part, to
model the USEPA field demonstration and help understand  more
fully the behavior observed during that demonstration.
Theoretical Basis

  The theoretical basis for predicting the performance of the Block
Displacement Process  is  founded  in previous work applied to
hydraulic fracture of rock formations. This earlier work was used
to predict pumping pressures  and flows as well as fracture orienta-
tion for oil recovery enhancement. The interaction  of multiple,
horizontally oriented fractures has been discussed  by several in-
vestigators.2
  A rigorous mathematical model has been developed' for predict-
ing both fluid slurry properties and fracture propagation geometry
for single and multiple interactive fractures. This model employs a
surface integral approach utilizing dislocation theory to predict
fracture behavior. Parametric inputs include slurry properties such
as viscosity  and pumping  rate, and  geologic properties including
modulus  of  elasticity,  poisson  ratio, permeability  and in  situ
stresses.
  The work  described  in  this paper compared the theoretically
predicted behavior to that observed in the  one-tenth scale model
tests. Performance observed during these model tests was, in  turn,
compared to that  observed during the full-scale field  demonstra-
tion.

EXPERIMENTAL FACILITIES
AND EQUIPMENT
Test Pit

  The testing program  was  performed  in  the materials testing
laboratory at FMI. A pit 7 ft  wide,  12 ft long and 6 ft deep in the
shop floor allowed for approximately one-tenth scale evaluation of
the previous Florida demonstration. All testing was conducted in a
uniformly graded concrete sand.

Instrumentation

  Testing was monitored with  a wide range of standard and custom
built instrumentation.
  Bentonite slurry properties were measured with a  Marsh funnel,
Shearometer, Fann filter press and unit weight bucket using  stan-
dard  API procedures.  Slurry  pressures  during pumping  were
measured with standard pressure gauges.
  In situ soil stresses were monitored using  transducers developed
specifically  for  this project. These  were  uniaxial,  oil   filled
diaphragm type units designed to operate at pressure levels from 0
to 5  psi. These devices were  used throughout the  program to
measure both horizontal and vertical stresses.
  The need to determine  the vertical  displacement of the soil at
several depths demanded the development of a simple yet accurate
vertical displacement indicator.  It consisted of a 2 in. diameter steel
washer  with a 36 in. long, 1/16 in. diameter  wire welded perpen-
dicular  to the washer face. The washer was placed at the desired
depth during soil backfilling. The washer was held in a fixed posi-
tion  relative to  the  soil around it while  the wire slid  through the
overburden. The central hole in the washer allowed for the vertical
stacking of displacement gauges at a single reference  point  in the
horizontal plane. Using gauges at several levels, the vertical plane at
which a fracture occurred was easily determined during testing.

Testing Procedures
  During the full-scale demonstration of block displacement in
Florida, a soil/bentonite slurry  was pumped into drilled and cased
holes that extended down 23 ft. These holes terminated in pancake
shaped  notches approximately  6  ft in diameter, produced by a
horizontal hydraulic jet, rotated from the surface to erode a notch
at the base of the drilled hole.
  The  cased  boreholes  and  notches  were simulated  in  the
laboratory by  tubes from the  surface  and  specially  configured
notches placed during the preparation and compaction of the test
bed.  Thin  metal  discs covering  7 in. diameter gravel   packs
simulated the notch.  This notch  provided the reservoir of high
pressure slurry from which the  fracture developed.
  For each test conducted,  a specific test bed configuration was
prepared, bentonite slurry  was pumped into the tubes and notches,
and  pressures,  flows and  soil displacements were  monitored and
recorded. Depending on the plan, pumping continued until either a
specific slurry volume had been placed, a target displacement had
been reached or a fracture appeared at the surface.
  Following the completion of each test, the pit was excavated and
the resulting fracture network and seal configuration was observed.

TESTING

  The testing program consisted of seven tests (Table  1).
  The  first  four  tests were  preliminary  in nature, serving  to
shakedown test equipment and procedures and to identify the ade-
quacy and limitations of  the instrumentation. During these tests,
fracture  mechanisms  and  slurry  behavior  were  explored. Dif-
ferences between the characteristics of the 4% and 8% bentonite
slurries were documented. It was determined that the 4% slurry was
superior for initial  fracturing,  while the 8% slurry tended  to in-
crease the final thickness of the fractures.

                             Table 1
                     Summary of Test Program
 Test
        Test Objection
                        Test Specification*
        Study muni fracture   Single well, 15 in. deep; Horizontal fracture was
        behavior, lest laboratory 1.5 gal of 4% slurry    formed, leak-off zone
        procedure!                           is day Tilled
        Study initial fracture   Single well, 30 in. deep; Permanent change in
        behavior, refine pro-   2 gal of 8*1 slurry     horizontal stress, conical
        cedurra                             and horizontal fractures
                                         found
        Study fracturing to the Single well, 20 in. deep; 3.5 gal of slurry
        ground surface. Com*  8*V slurry           pumped, fracture slope
        pare the effects of 4
-------
  The designs of Test Nos. 5^ an* 7 were based on the results of
the earlier tests. The significant aspects of these tests are presented
hi the following paragraphs.

Single Point Well Test

  The purpose of Test No. 5 was to maximize the extent, uniform-
ity and thickness of a slurry-filled fracture produced from a single
well point. A single notch was placed 30 in. below grade.
  Based on the results of previous testing, fracturing was initiated
with 5 gal of 4% bentonite slurry. After the initial fracturing, an
8% bentonite slurry was pumped to increase the vertical displace-
ment. Black concrete dye was added periodically to the slurry dur-
ing the pumping process to help  identify the fracture zones during
excavation.
  The soil surface was displaced upward during this test, taking a
domed  profile with the largest displacement centered over the
noich. This dome was over 5 ft in diameter and had a maximum
vertical displacement of 0.3 in.  at the center. Two feet from the
center, the average displacement was 0.1 in.
  Excavation of Test No. 5 revealed a bowl shaped surface approx-
imately 3 ft in diameter. This surface was the top of the leak-off
zone  above the fracture. When this surface was excavated by cross
sectioning,  it could be seen that the top  of the leak-off zone
paralleled  closely  the  actual fracture surface. One  significant
feature was the divergence of a  single fracture into several, as the
fracture surface extended radially from the notch. A second feature
of the fractures that was revealed in the cross sectioning was the
development of a dominant fracture. Many small fractures were
present initially. However, as the thickened, lifting slurry was in-
jected, a single fracture became dominant, causing all other frac-
tures to arrest.
  To test for the%ontinuity of the seal, the "bowl" formed by Test
No. 5 was filled with water to a  depth of 4 in. The water level was
monitored for three days. This test indicated that the seal was con-
tinuous. When the test  block was cross-sectioned, the major frac-
ture,  in fact, was not found to be continuous over the entire sur-
face. ^Rather, the  clay  permeated  leak-off zones of unconnected
fractures intersected, producing a continuous,  low permeability
barrier.
  The most significant aspect of this test was the apparent con-
tinuity of the seal formed. Based on the results of this test, the leak-
off zone could have a major effect on the overall permeability of
the seal.
  The 4% fracturing slurry followed by 8% displacement slurry ap-
peared to work well as a system. This may be caused by the leak-off
zone of the thin slurry generating an effect on the in situ stress field
and soil permeability.
  A major question in fracture mechanics is posed by this test. "As
the fractures  grow larger in thickness, do  adjacent fractures
coalesce or diverge?" Test No. 6, a three well point test, was
designed to study the interactions of fractures.

Three Point Well Test
  The purpose of Test No. 6 was to study the interaction of frac-
tures propagating  from three distinct injection points. Of primary
interest was the behavior of the fracturing within the zone sur-
rounded by the injection points. The test was intended to generate
fracture seams between 0.25 in.  and 1 in. in thickness. The three
well points were set up in a triangular pattern,  24 in. from each
other. Black cement dye was again added to the slurry to aid in
discrimination during excavation.
  The slurry injection process of Test No. 6 was conducted over a
2.25  hr period during  which a total of 25.5 gal of slurry was
pumped. The test ended when the slurry fracture broke to the sand
surface.
  The test began by pumping 7.5 gal of 4% bentonite fracturing
slurry. This slurry was  pumped sequentially to each of the three
well poults for 1 min each to assure that an equal volume of slurry
was pumped to each well point.
                           Figure 4
             Photograph of Excavated Test Block No. 6
                           Figure 5
                   Interaction of Two Fractures
  After 7.5 gal  of fracturing slurry were pumped, the mix was
changed  to the  8% bentonite lifting  slurry.  This mixture was
pumped at 30 sec intervals, sequentially to each well point.
  All of the displacement gauges indicated some lift during the test.
The average displacement of the surface gauges was about 0.4 in.
The maximum surface displacement was 0.48 in.
  Several days  after pumping, the overburden of sand was re-
moved from the test pit, exposing the surface of the leak-off zone.
Figure 4 is a photograph of this surface. The string grid evident in
the fiture is on 2 ft centers at the level of the original ground sur-
face.
  The three injection wells, holes left by core sampling, a wire
gauge and the fracture outcrop point can be seen in Figure 4. The
block is approximately 6 ft by 5 ft and roughly circular.
  A number of offsets can be seen in the fracture surface in Figure
4. The offset indicates that one fracture seam is passing under a se-
cond  fracture, and that two  fractures  have not coalesced into a
single seam. When the block was cross-sectioned, the interaction of
two fractures could be seen much more closely.
  An excellent example of fracture behavior is shown in Figure 5.
As the fractures approach each other, they tend to  diverge, one go-
ing up, the other down. After the fractures propagate past each
other, they then tend to converge. Eventually the fractures connect
and one leg continues growing while the other arrests.
  The fracture seams  were thinner and less well-developed at
greater distances from the notch. Because of this, the interaction of
fractures could be studied chronologically. By looking at two frac-
tures at the block rim, and then excavating through them toward
                                                                                           BARRIER TECHNOLOGY       137

-------
the center  of the block, the development of a fracture could be
observed.
  The fractures that were of greatest interest were those generated
between the  three well points. It can be seen in Figure 4 that the
block surface is fairly level between the well points. The fractures
appear to have a positive influence on each other in this area. When
excavated, the  fractures were fairly flat and fully interconnected
forming a  single fracture plane.
  The most significant result of  this test was the demonstration of
the interaction between  fractures. The test  demonstrated that in-
dependent fractures do tend to coalesce and  form  a single fracture.
The test also indicated  that fractures started  from  independent
sources will also coalesce, and that these fractures have a positive
influence on  each other. The fractures between the three well points
tended to be flatter and thicker than those  beyond the well point
boundary.
Seven Well Point Block  Displacement

  The purpose of this test was to demonstrate the key features of
the block displacement process.  Test No. 7 was a  1/10 scale model
of the geometry of the USEPA-Florida, Block Displacement test
site. The plan was to create a set of fractures from seven injection
wells, to consolidate these fractures into a  single continuous clay
layer and  then displace  the block 1 in.  vertically in  a  controlled
fashion.
   In this test, an artificial perimeter was constructed to define the
sides of the block,  much as a slurry wall would in a full-scale lift.
   The test was further intended to demonstrate  that once an ar-
tificial geologic layer has been created it can be  remobilized, and
-the block further displaced upward in a well-controlled fashion.
   The last major purpose of the test was to demonstrate the con-
tinuity of the clay layer by conducting a leak  test. From this test, an
effective permeability could be calculated.
  The laboratory setup is shown in Figure 6. This photograph was
taken after the block had been displaced 1 in. vertically.
  There were seven injection wells in this test. The wells were 27 in.
apart, forming a hexagonal pattern about the central well. At the
base of each  well, 30 in. below grade, was a 7 in. diameter artificial
notch.
   An artificial  perimeter was constructed for this  test  to define the
block  size  and  to control the extent of fracturing. This perimeter
was made of a plastic sheet placed in the soil.  The perimeter was 6 ft
in diameter at the ground surface and tapered inward  1.5 in. at the
perimeter to  a 5.75 ft diameter base at a depth of 30  in.
  The soil  displacement was measured with 18 dial gauges mounted
on  a stationary  steel  frame. These gauges allowed continuous
monitoring of vertical displacement during the test.
  The initial test consisted of fracturing and lifting the block 1 in.
vertically. This was done by pumping 35 gal of mud during a 2 hr
 period. After the initial test, the block was reactivated and lifted in
 small increments periodically over several weeks. Finally, after the
 vertical displacement testing was completed, a leak test was per-
 formed to assess the effective permeability of the installed seal.
   The first step in block displacement is to hydraulicaJly fracture a
 basal plane under the soil to be  lifted. From the previous tests, it
 was determined that a thin slurry was the most effective fracturing
 fluid. In Test No. 7, 8.5 gal of 4% bentonite slurry were used as the
 fracturing fluid. As this slurry was injected,  the pumping pressure
 slowly rose to 7 to  10 psi at the notch, and then quickly dropoped
 off to 2 psi as the soil fractured. This process, the pressure slowly
 rising and then  quickly dropping off, repeated itself as the block
 fractured.
   After a basal  plane was fractured under the block, this fracture
 was expanded vertically to displace the block. An 8% bentonite
 slurry was used  as a displacement mud. The pumping pressure in-
 itially rose to 15 psi. After pumping 19  gal of lifting slurry, the
 pumping pressure had dropped to 3 psi. This  is approximately 1 psi
 above overburden pressure.
   The slurry was pumped sequentially to each injection well. The
 vertical displacement of the block determined the amount of slurry
 pumped  to each well. A dial gauge recorded the vertical displace-
 ment at each well point. During the test, each well point was lifted
 sequentially 0.025 in.  Over the 6  ft diameter  of the block, the sur-
 face could be held level within  ± 0.005 in. during the entire block
 lift.  Scaling  this to a full-scale  lift, each well was lifted in in-
 crements of 0.25 in., and the block was held level within 0.063 in.
                            Figure 7
           Leakage Resistance, Permeability/Seal Thickness
                           Figure 6
     Laboratory Demonstration of Block Displacement Test No. 7
  After the first day of testing, 35 gal of slurry had been pumped
and the block had been lifted 1.04 in. Over the next 15 hr., with no
pumping, the block settled 0.48 in. This settling  was caused by
water leaking from the bentonite slurry. In practice, this settling
would be much less due to the structural bulking of the soil compo-
nent of a soil bentonite slurry. In these  1/10 scale tests, no soil is
added to the slurry during lift.
  On the second day of testing, the block was relifted. When slurry
pumping resumed, the pump pressure peaked at 9.5 psi and then
broke back to 3 psi. This was a typical pattern throughout the
subsequent lifts. It required 7.5 to 10.5  psi to reactivate the fracture
and 2.5  to 3.5 psi to lift the block. This is 0.5 to 1.5 psi greater than
the overburden pressure.
  The block was lifted a total of 2.52 in., settled 1.59 in. and had a
net vertical displacement of 0.93 in. The maximum displacement
was 1.04 in., on the first day of testing.  The maximum settlement
was 0.48 in., between the first and second test days.
  A  total of 44 gal of slurry were pumped during the test; 8 gal of
4% bentonite fracturing slurry, and  36 gal of 8%  bentonite lifting
slurry.
138
          BARRIER TECHNOLOGY

-------
  At the conclusion of the block displacement test, a leak test WSL,
conducted to determine the average permeability of the bentonite
  This leak test was the analog of a field draw down test. In a field
situation, a draw down test is the only practical test of seal con-
tinuity. The seal cannot be excavated and examined as it can in the
laboratory environment.  The block,  as defined by the plastic
perimeter and basal displacement fracture, was filled with water.
The water level was observed over the next 96 hr (Fig. 7). Above the
graph  is the average leakage  resistance  as  defined  by k/t,
permeability/seal thickness.
  During excavation, the entire overburden of sand and the plastic
sheet perimeter were removed. The exposed leak-off zone surface,
the block perimeter and  two exploration trenches are shown in
Figure 8. The shape of this fracture surface is roughly the intersec-
tion of seven bowls, each bowl centered about a well point.
  The six outer bowls  have been truncated  by the  perimeter.  In
general,  the leak-off zone surface parallels the clay fracture seam.
The leak-off zone does tend to be thicker above the notches.
  A cross-section through the block is shown in Figure 9. The scale
below the  central well is  marked in 1 in. increments. The  major
feature of this cross-section is the continuous fracture seam, ap-
proximately  1 in.  thick, traversing the entire block. The leak-off
zone, especially around the central well, is also clearly visible in the
photograph.
   Two other features can also be seen.  A number of small fractures
are visible above and below the main fracture. It is believed that
                            Figure 8
                     Excavation of Test No. 7
                            Figure 9
  Photograph of Cross Section Six Inches North of Central Well Point,
                           Test No. 7
                           Figure 10
Photograph of Cross Section Fifteen Inches South of Central Well Point,
                          Test No. 7

these fractures developed concurrently with the main fracture. As
the main fracture developed, the development of these secondary
fractures was arrested. The central bowl shape can also be seen
clearly in this cross-section.
  The  typical "inverted top hat" configuration of the fractures
around the notches is shown in Figure 10. At the notch edge, the
fracture tends to rise vertically for a few inches and then abruptly
turn horizontal. This is due to a  local shear feature in the soil,
caused by the structural rigidity of the steel notch cap. This result
seems to be a flaw in the modeling procedure; however, this localiz-
ed shear failure appears to be confined to the notch area with little
overall effect.  It can  be  seen in the photograph that a  single
horizontal fracture traverses the block between the notches.
  The  most significant feature of this test was the well controlled
lifting  of the test  block. This test demonstrated that a  single
horizontal fracture surface could be produced under a block of soil
and then widened by lifting the block vertically in a controlled man-
ner. The test also demonstrated that fractures emanating from in-
dependent wells will influence each other in a positive way, and,
given sufficient vertical displacement  and horizontal extent, frac-
tures will coalesce into a single plane.
  Test No. 7 was the first test in which a leakage test could be per-
formed. The positive result  is significant for future field applica-
tions as a draw down test would be the proof of a complete barrier.
For this laboratory test, the leakage rate  out of the block was
measured rather than the groundwater infiltration rate, as might be
done in the  field.
  Test No. 7 provided an initial exploration of fracture mechanics
in unconsolidated soils. This behavior must be further evaluated
for varying  conditions. One of the most critical variables is the
groundwater table. Test  No. 7 was run above the groundwater
table. Other parameters such as soil grain size, soil layering and soil
stress must be evaluated.


CONCLUSIONS

  The  tests conducted to date have indicated that fracture interac-
tive processes and displacement control processes can be studied
for specific  geologic conditions. The fracture initiation behavior
from each disc shaped notch, however, does not appear to be ade-
quately modeled. The influence of the simulated notch and/or the
use of  nonscaled soil grain size may be contributing to this  inade-
quacy.
  The  apparent deviation from predicted full-scale behavior is a
local steeply rising shear fracture which then turns horizontal and
extends from there as predicted.  This discrepancy has not had
significant impact on the  overall test results as it extends over a
distance of only 2 to 3 in. from the tip of the simulated notches.
                                                                                             BARRIER TECHNOLOGY
                                                           139

-------
   The  prediction that  the  fractures remain horizontal  over a
 substantial distance and the prediction that the influence of the free
 surface will dominate over in situ stress and viscosity effects has
 been demonstrated. The favorable influence of slurry leak off on
 local stresses has been shown to be significant in keeping the frac-
 tures horizontal by locally increasing the horizontal stress.
   Comparisons between the performance of the 1/10 scale model
 tests and observations in the field demonstration are limited.  To
 date, the model tests have not been performed under fully saturated
 soil conditions, whereas the field demonstration was performed in
 saturated sand with a groundwater table only a few feet below the
 surface. The  scaled  pumping pressure and  lift geometry  appear
 similar. The field demonstration  was hindered by unsatisfactory
 perimeter  conditions,  prohibiting  a  complete  comparison.
 However, local   surficial  soil deformation  and  displacement
 behavior were similar for the full  and 1/10 scaled tests.
   The  1/10 scale facility can be used for future testing to evaluate
 the influence of groundwater as well as for modeling other geologic
 conditions including stratification, discontinuties, large grain size
 material such as boulders or cobbles and irregularities in surface
 topography. The influence of perimeter construction technique and
 construction sequencing can also be evaluated.
   The field testing of the block displacement process carried out
 under funding from the USEPA and the laboratory studies carried
 out under a grant from the National Science Foundation show that
 block displacement is  a  viable technique  both theoretically and
 practically for introducing horizontal low permeability layers into
 unconsolidated soils.
   It is believed that block displacement can be a powerful tool in
 the control  of hazardous waste.
REFERENCES

I. Brunsing, T.P. and Grube, W.E., Jr., "A Block Displacement Tech-
   nique to Isolate Uncontrolled Hazardous Waste Sites." Proc. National
   Conference on Management of Uncontrolled Hazardous Waste Sites,
   Washington, DC, Nov. 1982.
2. Huck, P.J., el al., "Innovative Geotechnical  Approaches  to the Re-
   medial  In Situ Treatment of Hazardous Materials Disposal  Sites,"
   Proc. of the 1990 National Conference on Hazardous Materials Spills,
   May 1980, Louisville, KY, 421-426.
3. Narendran,  V.M. and  Cleary, M.P.,  "Elastostalic  Interaction of
   Multiple Arbitrarily Shaped Cracks in Plane Inhomogeneous Regions,"
   Engineering Fracture Mechanics, 19,  1984, 481-506.
140
         BARRIER TECHNOLOGY

-------
      EXPLOSIVES WASTE DISPOSAL SITES: A DOD-WIDE
                                PROBLEM  CASE STUDY:
          Milan Army Ammunition Plant O-Line  Settling Ponds

                                            PETER K. WIRTH
                           U.S.  Army Toxic and Hazardous Materials Agency
                                            Assessments Division
                                  Aberdeen Proving Ground,  Maryland
INTRODUCTION

  The disposal  of explosives-laden waste water from munitions
manufacturing prior to implementation of currently acceptable
environmental controls (e.g., filtration, activated carbon absorp-
tion) commonly entailed the use of earthen surface impoundments
(ponds) in conjunction with drainage ditches. This procedure re-
sulted in contamination of surface add groundwater and asso-
ciated soils and sediments.  To prevent additional environmental
damage, remedial actions must be undertaken. These actions must
be based upon site investigations and  proper planning/design of
the appropriate remedial cleanup measures.
  In this paper, the author describes the site investigation and re-
medial action (under construction) at the 0-line settling ponds lo-
cated at Milan Army Ammunition Plant (MAAP), Milan, Ten-
nessee.  This project has been implemented as part of the U.S.
Army Installation Restoration Program through  the U.S. Army
Toxic and Hazardous Materials Agency located at Aberdeen Prov-
ing Ground  (Edgewood Area),  Maryland.  The assistance of
MAAP, the U.S.  Army Engineer Division, Huntsville,  and the
U.S. Army Engineer District, Mobile,  has been  instrumental in
the progress of this action and is greatly appreciated. •
           1X1000    XUOOO     30000     UUOO    353000     3MOOO
                           trr»  cotwniNHTES

                        Figure 1
Location of City of Milan, Milan Army Ammunition Plant, and the 0-line
                      Settling Ponds
 SITE DESCRIPTION AND HISTORY

   The 12-acre 0-line settling ponds sites is located within MAAF,
 approximately 5 miles east of the City of Milan, Tennessee (Fig. 1).
 The ponds are part of the 0-line facility that is used for conven-
 tional munition demilitarization. Defective and outdated munitions
 loaded, assembled and packed (LAP) at the plant were disposed
 at this line.
   The major function of 0-line was to remove explosives (2, 4,
 6-trinitrotoluene (TNT) and cyclotrimethylenetrinitramine (RDX)
 from bombs and projectiles by injecting a high pressure stream of
 hot water and steam into the open cavity of the munitions. Waste
 explosives were separated  from the resulting water phase  and
 collected for reuse or for disposal at the MAAP explosive burn-
 ing grounds. Effluent washwater was then passed through baffled
 concrete sumps outside the wash-out building where it was cooled
 and entrained explosives particles were removed by screens. Cool-
 ing was aided by a cold water spray at the surface of the sump
 chambers. The screens and sumps were periodically cleaned to re-
 move collected explosive particles.
   Until 1941, the water effluent from the sumps was discharged to
 an open drainage which ran through  the 0-line area. In 1941, hold-
 ing ponds were constructed at the site to provide an additional
 settling capacity for the waste water.  The ponds consist of 11 indi-
 vidual basins connected by spillways  and open ditches with baffles
 and distribution boxes to allow several configurations of ponds to
 be employed in series (Fig. 2). The ponds have a total capacity of
 approximately 5,500,000 gal and cover an area of about 280,000
 ft2 (excluding the dikes).
   In operation the ponds received  water from the plant sump
 through an open concrete flume. Most of the solid explosive par-
 ticles settled to the bottom  of the first receiving basin. Effluent
 from the last basin in the series overflowed through a bank of
 carbon-filled tanks before being discharged to the area drainage
 ditch. The carbon from the tanks was periodically removed and
 burned. The drainage ditch ultimately discharged across the north
 boundary of the installation to the Rutherford Fork of the Obion
 River.
  In 1971, sediments were dredged from the ponds using a drag
line, and the dredged spoils were placed at the northwest corner of
the pond area. An attempt was made to burn the sediments at the
explosives burning ground; since  the  material would not burn  the
remaining dredged spoils were left in the area.              '
  In 1981, MAAP drained the ponds, treated the effluent  moved
the spoils pile back into the dredged ponds and lined the empty

  CONTAMINATED GROUNDWATER CONTROL     141

-------
                            Figure 2
            0-line Settling Ponds with Spillways and Baffles
                                              r^a
                                              uSL.-.tvr'CijJa?.-

                           Figure 3
       Geologic Cross Section for Milan Army Ammunition Plant
ponds with synthetic liners as a temporary remedial measure  to
prevent additional ground water contamination.
SITE INVESTIGATION
  Field investigations were conducted  at MAAP and  the 0-line
area in order to define the magnitude of contamination resulting
from the 0-line operation. To evaluate the impact  of  the 0-line
settling ponds site on the environment, production records were re-
viewed to assess the amounts and types of waste disposed at the
site. In addition, geohydrologic information, topographical  data
and  meteorological data pertinent  to the area were reviewed  to
determine probable pathways of  contaminant migration (e.g.,
246-TNT and RDX) from the site.
  Results of this review determined that explosives contamination
migration was probable via two mechanisms: (1)  surface water
migration of explosives by means of the drainage ditch  located  in
and adjacent to the 0-line area; and (2) groundwater migration  of
explosives by means of the settling ponds and, to a degree, the
drainage ditch. Both the settling ponds and drainage ditch allowed
for infiltration of explosives into the groundwater flow system be-
neath MAAP when waste  water was present. The extent of  con-
tamination from the site could  not be assessed from the review.
Actual installation  of monitor  wells and collection of environ-
mental samples from the area for chemical analyses was required.
                                                          Monitor well installations were based on the geohydrologic set-
                                                        ting present at 0-line. Available  information  indicated the Clai-
                                                        borne and Wilcox formations as the shallow water bearing units
                                                        (aquifer) underlying MAAP. These formations consist chiefly of
                                                        sands with lenses  and interbeds  of clay at various  stratigraphic
                                                        horizons with an average total depth  of 300 ft beneath  MAAP
                                                        (Fig. 3). Groundwater flow at the site was thought to be in a north-
                                                        west (NW) direction, and this was close to the actual situation at
                                                        the 0-line site. However, the direction of groundwater  flow at
                                                        MAAP  is greatly influenced by topography and surface  streams
                                                        that alter  flow patterns throughout the plant (Fig. 4). Depths of
                                                        the wells varied to monitor the upper,  middle and lower portions
                                                        of the aquifer for explosives contamination. The correct placement
                                                        of monitor wells was essential in evaluating the magnitude  of con-
                                                        tamination originating from the site.  In this case, monitor wells
                                                        were placed downgradient of groundwater flow  from the  settling
                                                        ponds within distances where contamination was anticipated to be
                                                        present based upon groundwater hydraulics of the area.
                                                          Water samples were collected from the drainage ditch,  settling
                                                        ponds and groundwater monitor wells. Analysis  was done for ex-
                                                        plosive compounds and associated degradation products (Table 1).
                                                        In  addition, geohydrologic  information was  collected from toe
                                                        monitor wells  (Fig.  5) to define groundwater flow, soil types and
                                                        groundwater hydraulic properties specific to the site.
                                                          Results of the sampling indicated the 0-line settling ponds and
                                                        associated  groundwater as major areas of contamination. Sedi-
                                                        ments from the ponds indicated up  to  approximately  5ft ex-
                                                        plosives content in the top 12 in. of materials (Fig. 6). These con-
                                                        centrations decreased with increasing depth of sediment indicating
                                                        leaching of contaminants into the subsurface zones. Groundwater
                                                        was contaminated with explosives 1.25 miles downgradient of the
                                                        settling ponds (Figs. 7, 8 and 9) and flowing in the  north-north-
                                                        west (NNW)  direction.  The contaminated  groundwater zone
                                                        (plume) was limited to the middle section of the Claiborne Forma-
                                                        tion indicating vertical stratification of the contaminants. Some
                                                        levels of contaminations were above U.S. Army Interim  Drink-
                                                        ing Water Criteria (RDX-34 Mg/l and TNT-44 ^g/l).
                                                          The presence of explosives at high  levels (Fig. 10) adjacent to
                                                        the settling ponds and the existence of  a sizeable plume migrating
                                                        toward the installation boundary adjacent to the Rutherford Fork
                                                        of the Obion  River indicated a need for remedial action  to pre-
                                                        vent further environmental damage. Samples collected in the drain-
                                                          " 19)1000 •
                                                            1M70OO
                                                                113000
                                                                                     M1000      MOOOO

                                                                                      UTM COORDINATES
                                                                                     Figure 4
                                                             Piezometric Surface of Groundwater in the Claiborne Formation
142
CONTAMINATED GROUNDWATER CONTROL

-------
                             Table 1
      Explosive Compounds and Associated Degradation Products
              2,4, 6-Trinitrotoluene (246-TNT)
              1, 3, 5-Trinitrobenzene (135-TNB)
              2,4-Dinitrotoluene (24-DNT)
              2, 6-Dinitrotoluene (26-DNT)
              1, 3-Dinitrobenzene(13-DNB)
              Cyclotrimethylenetrinitramine(RDX)
              Cyclotetramethylene Tetranitramine (HMX)


age ditch indicated explosives were present, but the low levels
found (under 50.0  mg/kg) did not justify additional study or re-
medial actions.
REMEDIAL ACTION

  Various remedial action (closure) alternatives  were considered
for use at the 0-line  settling ponds site. These include:
•In-place containment using  migration barriers such as contain-
 ment walls and low-permeability caps
•On-site treatment of sediments using rotary kiln incineration
•On-site waste disposal in a newly developed facility
•Removal and off-site disposal/treatment
  The selection of a remedial action was largely restricted to the
in-place  containment option due to restrictions on disposal of re-
active wastes (i.e., explosives) into landfills and the lack of proven
technologies (e.g., incineration) for treatment. Treatment technol-
ogies are currently  being developed and should be available with-
in the near  future  for treatment of explosives waste. These tech-
nologies are needed for sites where an in-place containment action
is not suitable as a means of closure.
  The 0-line site was very  favorable to  an in-place containment
closure.  The geohydrologic conditions at the site provide for  ade-
quate isolation of the waste materials after installation of the low
permeable cover system (grass/topsoil/clay-gravel-clay)  and con-
tainment wall. The depth of groundwater below  the ponds,  sur-
                       CONCENTRATION  (mg/kg)
                   10,000      20,000     30,000     40,000
                             Figure 5
    Monitor Wells Located in the Vicinity of the 0-line Settling Ponds
      (Reprint with permission from USGS Quadrangle, Atwood, Tennessee)
    12
ti   18
    24
    30
    36
                                                                                           I
                                                                                                       I
                                                                                                                  I
                                               A   2,4,6-TNT

                                               • -       RDX
                            Figure 6
          0-line Settling Pond 3-Inlet; Contamination Profile
                                                                                 9 - MONITOR WELL   GROUNDVIATEH DATA - ufl/l

                                                                                 T - MONITOR WELL
                                                                                                 Figure 7
                                                                             Groundwater Contamination in the Upper Section of the
                                                                                           -Claiborne Formation
                                                                                       - MONITOR WELL    GROUNDWATER DAT* - n

                                                                                      . - MONITOR *CLL
                            Figure 8
       Groundwater Contamination in the Middle Section of the
                       Claiborne Formation
                                                                       CONTAMINATED GROUNDWATER CONTROL       143

-------
                            Figure 9
        Groundwater Coniamination in the Lower Section of the
                       Claiborne Formation
                           Figure 10
        0-line Settling Ponds Groundwater Contamination Plume
face drainage and soil types are adequate to prevent surface water
and groundwater from contacting the contaminated sediments and
forming leachate that could flow into the groundwater system be-
neath the ponds. In addition,  borrow material (e.g., clay, inert
fill used for construction) is available at MAAP
   The closure at 0-line (Fig. 11) acts primarily as a diversion for
surface  water  from contacting  the contaminated  material in  the
ponds. A cross-section of the cover system (Fig. 12) illustrates the
method by which water is diverted. The system utilizes a proper
grade that allows a large portion of precipitation to runoff the site
or be removed through evapotranspiration of the grass cover. Any
remaining portions of water percolate through the upper soil into
a gravel drain  layer that allows for additional runoff. The gravel
layer contains a perimeter piping system that routes collected water
to the outside of the cover system.
  The clay layer is the final protective layer in the system. This
layer is designed to prevent percolation of water  for  an extended
period. By compaction of low permeability clays, percolation of
water is prevented until residual water in  the drain  layer causes
saturation. However, the  rate  of percolation through  the layer
after saturation is  minimal,  restricting  the  flow of leachate into
the groundwater flow system. Any possibility of lateral movement
of infiltrating  precipitation into the ponds  is prevented by  the
perimeter containment wall obstructing flow toward the ponds.
  Adequate  depth to groundwater below  the ponds is required
when using an in-place containment system. The bottom of  the
ponds should be a distance far enough from the groundwater sur-
face and capillary fringe area so that contaminated sediments are
not in  contact with  groundwater.  Otherwise, contaminants will
                                                         leach and  migrate  into the groundwater flow system.  Ground-
                                                         water depth at the 0-line site is approximately 40 ft below the bot-
                                                         tom  of the ponds. This depth is more than adequate for the con-
                                                         tainment system.
                                                           Construction of the containment system required an  assessment
                                                         of the explosive potential  of the sediments prior  to  any actual
                                                         earthmoving operations. The low content of explosives  in the sedi-
                                                         ments indicated a minimal potential for any ignition of explosives.
                                                         However, testing was  conducted to determine if ignition could
                                                         occur due to localized stresses on the sediments due to heavy earth-
                                                         moving equipment.
                                                           A  friction test using  the U.S.  Bureau of Mines Pendulum Fric-
                                                         tion  Apparatus was conducted  on sediment taken  from the site.
                                                         The sediment was mixed  with explosives at various levels ranging
                                                         from 0-25%. Results indicated that sediments with up to 20% ex-
                                                         plosives were insensitive to the testing procedure with no moisture
                                                         present in the sample. The presence of moisture and a maximum
                                                         concentration of  approximately 5% explosives in the 0-line site
                                                         sediments indicated earthmoving operations could be performed.
                                                           Construction is currently in progress at the 0-line site in MAAP.
                                                         Following completion in  the fall of  1984, a monitoring program
                                                         will be initiated to monitor groundwater for explosive compounds.
                                                         The effectiveness of the closure  will be evaluated based upon the
                                                         results of this monitoring.
                                                                                     Figure 11
                                                                      Profile of Recommended Cover Configuration
                                                             UC • liquid Tnni»li»ion Control    V
                                                             0! - Olvtnlon Inttrfjc*
                                                                                                   u - Slop* Anglo
                                                                                     Figure 12
                                                                     Liquid Routing Diagram of Cover Configuration
144
CONTAMINATED GROUNDWATER CONTROL

-------
                CASE HISTORY ORGANIC  RECOVERY  AND
                CONTAMINANT MIGRATION  SIMULATION

                                      PRESSLEY L. CAMPBELL, Ph.D.
                                D'Appolonia Waste Management Services Inc.
                                             Baton Rouge, Louisiana
INTRODUCTION

  A continuing concern associated with evaluation of hazardous
waste sites and groundwater contamination is the investigation,
identification and elimination of potential pathways of contami-
nant migration. In  this paper, the author describes a successful
project at a site which involved a myriad of horizontal and vertical
pathways. Migration of significant concentrations of organic con-
tamination, including free phase chemicals, occurred through pil-
ings, shallow permeable lenses and old well annular spaces. The
approaches  and techniques  which the  author employed, such as
computer modeling, can prove to be very useful in similar future
efforts.

BACKGROUND

  The site investigation and remedial evaluation discussed in this
paper were completed at a chemical plant site located in an indus-
trial district  adjacent to the Mississippi River. The plant, which
began operating prior to World War II, manufactures chemicals
including chlorinated organics such as ethylene dichloride (EDC).
Intermittment  spills, reboiler  cleaning operations  and leaking
pumps contributed EDC to a receiving stream located within the
plant. These discharges were discontinued or more closely con-
trolled after the early 1970s.
  Being heavier than water, the free phase organics accumulated
in ponding areas  located within the stream channel and migrated
vertically and horizontally through the underlying soil.
  Routine shallow RCRA groundwater monitoring revealed the
presence of dissolved EDC in some wells. Testing of plant water-
supply  wells indicated  that  EDC was also present in concentra-
tions as high as 20 mg/1 in limited portions of a major industrial
aquifer located at depths of about 250 to  400 ft below the plant
grade. Groundwater from this aquifer is used by the area plants for
cooling water.  Additional monitoring wells completed to depths
varying from 30 to  150 ft adjacent to the  wells of concern, con-
firmed the presence of subsurface free phase organics.
  During the workover of a plant active well, screened in the 400-
ft formation, free phase organics were detected leaking into the
well through a hole in the well casing and were recovered.
  The initial observed  pattern of subsurface contamination sug-
gested several possible migration pathways into the lower aquifer.

OBJECTIVES
  The purpose of the investigation by  D'Appolonia Waste Man-
agement Services was to assist plant personnel in determining the
pathways producing aquifer contamination and in implementing
 remedial measures. The specific objectives of these efforts were to:
 •Assist in managing and conducting a shallow groundwater con-
 tamination assessment and recovery program to define the mech-
 anism  by which organics were  contaminating the underlying
 aquifer and to remove contaminant sources
 •Simulate regional groundwater flow in the underlying aquifer to
 assess the potential for off-site migration and to maximize the
 containment and removal of the contaminant plume, regardless of
 changes in the pumpage of nearby wells at adjacent plants

COMBINED PROGRAM
  The investigation and recovery program were integrated into a
combined program in order to expedite control of the source of the
contamination and reduce program costs. The general approach of
the program is outlined below:
•Research historic plant operations, locations of former plant fa-
 cilities and review plant geotechnical boring data
•Plug all abandoned water-supply wells
•Seal the annular spaces in all operating water supply wells
•Complete borings to define the horizontal and vertical extent of
 the shallow contamination. Where free phase organics were en-
 countered, the borings were completed as recovery wells. Recov-
 ery wells were tested to facilitate design of the final recovery pro-
 gram and assess migration rates.
•Implement a continuing test  program for the approximately 15
 active water-supply wells and 30 recovery wells. Complete water-
 supply well testing to define aquifer characteristics.
•Compile and analyze regional aquifer information to define aqui-
 fer gradients and the boundary conditions for aquifer modeling
•Examine  chemical species data from water-supply wells  and re-
 covery  wells to identify potential sources, trends and effective-
 ness of remedial actions
•Simulate  regional groundwater flow for various aquifer pump-
 ing scenarios
•Simulate  contaminant mass transport and dispersion associated
 with alternative hypothetical sources and compare with the field-
 observed contaminant distribution
•Simulate alternative remedial pumping scenarios in order  to eval-
 uate long term aquifer cleanup
•Perform laboratory testing of artifically contaminated  aquifer
 media samples to assess retardation factors (attenuation) and sim-
 ulate long-term aquifer cleanup

SHALLOW CONTAMINATION ASSESSMENT

  During an eight month period, the plant engineering staff, sup-
ported by D'Appolonia, managed and implemented  a  shallow

  CONTAMINATED GROUNDWATER CONTROL      145

-------
drilling and recovery program to determine the mechanism of con-
taminant entry into the underlying aquifer and to remove the sus-
pected source(s). The objectives of this program included:
•Characterization of the geology of a complex depositional en-
 vironment
•Mapping of the migration pathways
•Estimation of the sources and  mass of shallow subsurface con-
 tamination
•Delineation of a cost effective method of contaminant removal
  At the time of paper preparation, 33 shallow piezometers (prev-
iously  installed), 30 recovery wells, four monitor  wells and ten test
borings have been completed at the site; the most recent are shown
in the general plant plan view (Fig. 1).
  The initial hypothesis was that contamination had somehow
reached the annular spaces of Wells W and X through the natural
drainage courses. Shallow subsurface cross-sections  were drawn,
pump  tests  of recovery wells were completed and  recovery well
yields were measured to determine the pathways for horizontal and
vertical migration. Soil sample head-space analysis with an organic
vapor  analyzer (in combination with subsequent  analytic  testing)
was used to map the extent of the contamination.
  With the construction of Recovery Wells J-O,  a major pathway
was identified. Completion of the cross-sections (Fig. 2),  along
with analysis of soil samples, revealed a continuous silt lens extend-
ing  below the plant site from approximately 40-65 ft  below grade.
Free phase contamination entered the lens through direct vertical
migration from a large  ponding area and by  moving  through
spaces around pilings under a nearby, abandoned foundation. Free
                                                        phase contamination then  migrated downgradient to the annular
                                                        spaces  of Wells W and  X. A similar phenomenon apparently
                                                        occurred in a separate plant area with resulting contamination mov-
                                                        ing to Well Z.
                                                          Borings in the vicinity of Wells Z, X and W revealed that the
                                                        contamination has moved down the annular spaces and thence hor-
                                                        izontally  through deeper permeable zones such as those at 120,
                                                        190 and 250  ft.  Shallow recovery wells were constructed through-
                                                        out the affected zone, and  deep recovery wells have been installed
                                                        near Wells W, X and Z.
                                                          The typical design of a recovery well is shown in Figure 3. Jet
                                                        pumps  with eductors were  utilized due to  the low well yields (less
                                                        than  5  gal/min) and the presence of free  phase organic solvents.
                                                        Early  operations  were  hampered due  to emulsification,  pump
                                                        breakdowns and problems in coordinating  maintenance work. Use
                                                        of polymers has reduced emulsification and, in turn, pump break-
                                                        down problems. On the average, each well is capable of produc-
                                                        ing 3000 gal/day. Each well is pumped 8 to 12 hr per day and then
                                                        allowed to recover.  Pump tests indicated that,  generally, a well
                                                        spacing of 100 ft would result  in maximum short  term recovery.
                                                        Some wells asre currently yielding as much as 5% free phase organ-
                                                        ics, a decrease of 20 to 30"% from the initial  well recoveries.
                                                          The recovered fluids are pumped to separation tanks and the free
                                                        phase organic portion is recycled  to the plant. The waste water is
                                                        transported to a stripper column and subsequently discharged  in
                                                        the plant wastewater system.
                                                          As contaminated fluids in the immediate vicinity of a recovery
                                                        well are withdrawn, the fluid is replaced by a combination of: (1)
                                                           Figure 1
                                       Chemical Plant Layout and Location of Monitoring Wells
146
CONTAMINATED GROUNDWATER CONTROL

-------
i  I
  I
I  "•
'  1
!  5
        lOUOMTALLT MOT TO KALE
  ICCfHO


O KCCOfCHt «LL

A TEST IOMHN

O MOMTOM «LL

O MODUCTIOM >CLL

S3 WELL »C*EEN IHT
                                                            Figure 2
                                             Geological Cross-Section Below the Plant Site
 heavier than water contaminants being drawn horizontally toward
 the well and (2) relatively clean water being drawn from the above
 more permeable portion of the silty zone. Consequently, recovery
 yields will decrease with time. The rate at which they will be re-
 moved is a function of the horizontal pore velocity. The horizontal
 pore velocity at a given distance is equal to permeability times the
 gradient at that distance.
   The permeability  of the silt zone varies widely.  The zone is  a
 distinct formation which probably has a permeability two to three
 orders  of magnitude higher  than  the  overlying and  underlying
 clays.  The zone itself, however, is heterogeneous, ranging from
 silty clay  to  medium grained sand. Permeabilities probably range
 from 10-' to 10-' cm/sec. This lithology is a result of a complex
 depositional environment. A degree of continuity for some of the
 clay, silt and sand lenses can be assumed. The underlying clay is
 relatively  continuous and has a tested permeability of about 1 X
 10-"cm/sec.
   Most of the contamination pooled in the sandy-silt lenses and
 subsequently migrated along pathways dictated by the slope of the
 underlying clay and the regional (southward) gradient.  The fact
 that the contaminants moved  from the ponding area  to Well X
 (1000 ft) in less than 20 years, at gradients of 1%, indicates that
 there is an average permeability, for some of the lenses, of at least
 10 ft/day (10-3 cm/sec).
   Most of the free phase contaminants will therefore be removed
 from the sandy lenses at fairly high velocities. Assuming an average
 permeability of 10 ~3 cm/sec for the sandy lenses, these lenses can
 be "cleaned" within a radius of 50 to 100 ft in one to three years.
     It was estimated that, during a period of up to three years, most of
     the additional free phase organics that can be recovered (about
     35% of total mass) will be removed from the adjacent silts.
       The recharge of clean water (for the long term removal of con-
     centrations) will flush the lenses to levels which pose even less risk
     of eventual migration into the underlying aquifers.


     WELL PLUGGING

        To reduce the risk of contamination to the underlying aquifers,
     the  annular spaces of all  abandoned  water-supply wells  in the
     potentially contaminated areas were plugged. Generally, the casing
     of these wells had been sealed with substantial surface and aquifer-
     level plugs. Only in recently constructed wells had the entire casing
     been grouted. As evident  from the intermittent contamination
     found at depths of about 120,  140 and 250 ft, organics were mi-
     grating from the upper source, down through the annular space
     along the less tortuous path  presented by the casing.
        Slant hole drilling into the annular space at a depth of about 70
     ft was completed and grout  injected until a surface return had been
     achieved.  Continued water  well pumping  in combination with the
     well plugging program has resulted in a 90% decrease in maximum
     aquifer  contaminant   concentrations.   Nonetheless,  computer
     modeling and field data demonstrate that  while the primary source
     has  been substantially reduced, the presence of contamination at
     120, 140 and 250 ft near the wells constitutes  a continuing source
     for migration around the grout until the recovery program is com-
     plete.
                                                                     CONTAMINATED GROUNDWATER CONTROL      147

-------
AQUIFER MODELING
  D'Appolonia emplolyed  GEOFLOW,  an in-house computer
software package, to simulate groundwater flow and contaminant
movement within the underlying aquifers. Contaminant concentra-
tions in the 250 to 400-ft aquifer approach 10 ppm EDC while
levels in the 500 to 600-ft aquifer are less than 100.ug/1 EDC. Thus,
contaminant migration modeling was focused on the 400-ft aquifer.
  GEOFLOW  is  a finite element  grid program which  idealizes
groundwater flow into a two-dimensional system. This assumption
is valid where variations in aquifer  thickness are much less than
total thickness. In the modeling, the aquifer was also assumed to
be homogeneous, isotropic and of infinite areal extent. Additional
data were developed through literature search and water-supply well
testing to define variations in aquifer thickness, permeability, stor-
age coefficient, effective porosity,  elevation and the retardation
(attenuation) factors of the aquifer  for the contaminants of con-
cern.
  The available data from agencies and nearby plant wells were
organized for input into the models of the 400- and 600-ft aquifers.
Interpretation was based on maps and tables which included well
and boring location maps, isopach  maps, structure maps (eleva-
tions of formations), potentiometric surface maps, geologic cross-
sections, water elevation versus time graphs, field permeability test
result tables and well pumping rate summaries.
  A finite element grid  system incorporating  the plant  and the
neighboring plants was generated. Each element of the grid system
represented a discrete  segment of  the  aquifer and  was  assigned
values  for aquifer thickness, permeability, storage coefficient,
effective porosity and elevation.
  Groundwater flow within the aquifers is primarily a function of
three parameters:
•Well locations
•Pumping rates
•The recharge (boundary conditions) at the periphery of the area
  modeled
  Sensitivity analyses were performed to assure that  the modeling
results  would be representative  of  real conditions. GEOFLOW
was utilized to solve the governing equations and thereby  simulate
flow and contaminant movement at the plant site. Initial modeling
confirmed the need to incorporate the complex geology of a greater
              SUMP
                                  CEMENT/BENTONlTE GROUT


                                   PROTECTIVE CASING
                                    BENTONITE PLUG
                                  HEAVY CONTAMINATION
                                GRAVEL -PACK
                                  HEAVY CONTAMINATION


                                 SCREEN, JOHNSON  IO-SLOT
                               Ł   WIRE WRAPPED
                               f
                                                                                                       J9M
                           Figure 3
                  Typical Organic Recovery Well
                                                                                  Figure 4
                                                                           Simulated Clean-Up Curve

                                                       area inclusive of the  hydrogeological regime below the Mississippi
                                                       River to simulate the hydrogeologic setting of the industrial area
                                                       surrounding the plant. The grid system  for later mass  transport
                                                       simulations was essentially the interior portion of the larger flow
                                                       grid system with constant flow boundaries.
                                                         Quality control and quality assurance checks of the input data
                                                       against  the raw data were performed to eliminate errors and to
                                                       assure that modeling was based on the best available data.
                                                         Several simulations, representing different sets of pumping rates,
                                                       were performed for both the 400 and 600-ft aquifers. The two crit-
                                                       ical scenarios were:
                                                       •Historical Case—Under historic operating conditions, regional
                                                        groundwater  flow within the 400-ft aquifer  is toward the plant
                                                        from all directions and is controlled by pumping.
                                                       • Worst Case for Plume Containment—The  worse  case analysis
                                                        predicted groundwater  movement through the plant to the south-
                                                        east in the unlikely event of the plant's pumping wells being totally
                                                       inactive and the pumping of nearby wells being at full capacity.
                                                         Three additional 400-ft aquifer simulations were performed to
                                                       illustrate the effects  of  alternative scenarios for the pumping of
                                                       nearby wells. Additional modeling for the 400-ft aquifer was com-
                                                       pleted to determine whether adjusting the distribution among plant
                                                       well pumpages would affect the  regional gradient toward the
                                                       plant site or the downward gradient between the shallow contam-
                                                       inated zones and  the 400-ft aquifer. These simulations were also
                                                       conducted to identify the optimal  location of a new production
                                                       well. The conclusions reached as a  result of the flow simulations
                                                       were:
                                                       •On-site contaminant plume control in the 400-ft aquifer can be
                                                        maintained even with significant reductions in pumping (50% re-
                                                        duction at Well Y; elimination of pumping at Well P).
                                                       •Under typical plant pumpage from the 400-ft aquifer, 1900 gal/
                                                        min, variations  in  pumpage distributions do not  significantly
                                                        affect the overall groundwater flow pattern toward the plant. The
                                                        primary change is limited to the area encompassed by the radius of
                                                        influence of the well.
                                                       •The vertical gradient is increased significantly only within  a few
                                                        feet of the pumping well. Accordingly, vertical migration is prob-
                                                        ably limited to near the  wells.
                                                         Similar conclusions were drawn  from the 600-ft aquifer simu-
                                                       lation. Based upon these findings, two general locations for a new
                                                       water supply well, Well  V, were recommended. In addition, Wells
                                                       W (now abandoned), X and perhaps Z were confirmed as poten-
                                                       tial vertical pathways to the underlying aquifers.

                                                       CONTAMINANT MIGRATION MODELING

                                                         GEOFLOW  was  used to  evaluate the potential  source(s) of
                                                       organics contamination  in the 400-ft aquifer. Potential combina-
148
CONTAMINATED GROUNDWATER CONTROL

-------
tions of contaminant source locations and historical pumping rates
that may have produced the existing distribution of organics in the
400-ft aquifer were evaluated to assess the likelihood of each par-
ticular source. The basis for determining the validity of a trans-
port simulation result, relative to existing conditions, was the de-
gree of correlation with water quality data for various wells (water
wells, recovery wells and observation wells) obtained during the
previous months. For example, transport modeling was undertaken
to test which of the two locations, Well W or X, was the more pre-
dominant source.
  The  results  from modeling a three year active leak were  com-
pared with the groundwater chemical analyses. A constant dimen-
sionless concentration  of 1.0 was used for the contamination
source.
  The simulation results illustrated the following features:
•For either source, the primary direction of the model plume move-
 ment was from the  source area toward Well Y. The plume also
 spread normal to the direction of  flow due to hydrodynamic dis-
 persion.
•For either source, Well Y intercepts most of the contamination.
•Once organics are introduced in the aquifer, their presence is ob-
 served at Well Y in a period of less than six months. The concen-
 trations reached approximate steady-state conditions within 1 to
 2 years.
  Calculations were made comparing the computed concentrations
to observed ones and  two conclusions became apparent:
•Well X was not the dominant source. Even though computed con-
 centrations at Well Y  were comparable to observed levels, the
 computed concentrations at  Well X were orders of magnitude
 higher than those actually observed.
•The computed ratios of concentrations for Wells Y and Z,  with
 W as the source, were consistent with the observed ratio.
  Subsequent additional field evidence supported the conclusion
that W was the dominant source. In particular, the simulated clean-
up curves at Well Y for a W source  are very much like the observed
data (Fig. 4).
  In  addition,  comparison  of contaminant constituent ratios
(EDC, perchloroethylene, trichloroethylene) for different recovery
wells snowed  that the ratios from Wells Y and  Z were compar-
able to  the ratios found in recovery wells near Well W and unlike
those for wells near Well X.

REMEDIAL ACTIONS
  The cleanup of the 400-ft  aquifer is a function  of the geograph-
ical dispersion of the higher concentrations of contamination and
of desorption rates. Contaminants primarily disperse longitudinally
with the gradient. However, they also disperse, to a lesser degree,
transverse to  the  gradient.  Cleanup occurs as contamination is
washed off of the soil grains, i.e.,  desorbed. The higher the  con-
centrations,  the longer the period  to clean up to a certain level.
The key then  to rapid cleanup is to prevent transverse dispersion
from increasing the areal extent of contamination. Consequently,
alternative pumping programs were evaluated to develop the most
effective pumping protocol.
  The amount of water required to  remove contamination from the
400-ft aquifer was the subject of  laboratory desorption tests by
D'Appolonia. A sand sample from the 400-ft aquifer was packed
to  a 32% porosity (approximately 20% effective porosity)  in a
column with a flexible wall.  The column was flushed with a solu-
tion containing 20 mg/1 EDC and lesser concentrations of other
chlorinated hydrocarbons.  Afterwards the column was  flushed
with distilled water and  the desorption rate for the various hydro-
carbons was observed.
  A 95% reduction in EDC concentration (from 20 ppm to 1 mg/1)
can be accomplished with 3.5 pore volumes of flushing. Estimates
involving the pore velocity between  Wells W and Y showed that
3.5 pore volumes  corresponded to  time periods of about  three
years. Since clean water does not actually flush the zones of higher
concentration, (the groundwater becomes more contaminated as it
moves toward the pumping well), the real time required to reduce
concentrations below 1 mg/1 could be slightly longer.
  It was concluded, through the analysis of modeling, pumpage
and chemistry data, that the contaminant plume is small and is
being rapidly removed at Well Y. These conclusions supported the
choice of the following remedial actions:
•Continued pumping of Well Y at its maximum rate
•Continued pumping of Wells Z, X, V and existing deep recovery
 wells
•Additional drilling and recovery from deeper zones near Well Z
 and annular "space" grouting
  Cleanup  of the 400-ft aquifer to concentrations less than 1  mg/1
was projected within 3 years. Even lower concentrations will de-
pend  on the  success of the  well annulus plugging program and
shallow organics recovery  program.  Analysis of future chemistry
and pumpage data, and possible further modeling, is important to
confirm and refine the above conclusions.
CONCLUSIONS
  The foregoing study demonstrates the efficiency of a combined
investigation/remedial program and the success of the approach
followed. A thorough hydrogeologic assessment, with painstaking
data gathering and careful analysis, was the crucial first step. Key
to recent  reductions in the contaminant concentrations in  the
underlying aquifers was  the quick  implementation  of remedial
measures to:
•Remove the source contamination (shallow recovery wells)
•Inhibit source migration (well plugging)
•Control plume migration (designed water supply well pumping
 program)
  Consequently, the plume  has been controlled before it has dis-
persed significantly, and an  adequate conceptual model of the sit-
uation has been developed. Neither would have been possible with-
out the timely input and interpretation of results from the recov-
ery well and modeling investigation program.
  The overall approach was successful because it represented  a
comprehensive cost-effective effort to collect and analyze readily
obtainable data in a timely manner. Of particular importance were:
•Development of an early understanding of regional aquifer char-
 acteristics
•Detailed analysis of previous plant geotechnical boring logs
•Field mapping of soil  contaminant levels with an organic vapor
 analyzer
•Organizing and  interpreting  chemical  analyses  and  pumping
 records for the recovery and water supply wells
•Formulating,  scrutinizing  and  refining hypotheses regarding
 potential sources and quantities  of contaminants, and migration
 pathways
•Mass transport/dispersion modeling as well as hydraulic modeling
 coupled with a flexible field program to develop calibrating and
 corroborating data
•Implementation of a management plan to concentrate efforts on
 the problem at hand,  i.e.,  generating a successful remedial pro-
 gram for  the plant and the aquifer
                                                                   CONTAMINATED GROUNDWATER CONTROL
                                                                                                                          149

-------
  DETERMINING CONTAMINANT MIGRATION  PATHWAYS
                               IN  FRACTURED BEDROCK

                                            PETER J. McGLEW
                                        J. ELLIOTT THOMAS, JR.
                                              Superfund Branch
                                  U.S. Environmental Protection  Agency
                                           Boston, Massachusetts
 INTRODUCTION
   The deep bedrock well which supplies water to an apartment
 complex was found to be contaminated during a hazardous waste
 site  investigation.  This  apartment  complex  (195 units) and 18
 homes are situated on 200 acres of developed land in  Southern
 New Hampshire.
   The investigation  revealed widespread contamination of the
 bedrock aquifer with volatile organic compounds. The  predomi-
 nant compounds found in the ground water include: 1,1-dichloro-
 ethane, trans   1,2-dichloroethylene, ethylbenzene, trichloroethy-
 lene and toluene.  The source area for the  contamination  that
 migrated to the apartment complex water supply well was approx-
 imately 1,500 ft away in a sparsely wooded area (Fig. 1).

 PHYSICAL SETTING

   The study area is  characterized by the low rolling hills of the
 glacial terrain common throughout southern New England.  The
 major drainage  basin for the area is the Merrimack  River. A
                        Figure 1
           Area of Oroundwater Contamination Study"

From USEPA Remedial Action Master Plan, NUS Corp., 1983


150      CONTAMINATED  OROUNDWATER CONTROL
tributary to the Merrimack is the surface drainage for the site. The
land surface of the site has a 4% slope to the southwest.
  The hills surrounding the area rise from 100 to 200 ft above the
valleys. Within a 1 mile radius of the site, the highest elevation is
460 ft above MSL and the lowest is 220 ft above MSL.
  The homes in the study area use septic tank leach field systems
for disposal of their household wastewater. Prior to the discovery
of the contamination  and the subsequent construction of a water
line, the residents did not have acce~ss to a municipal water supply.
Throughout  the area, the majority of residents rely  on deep
bedrock wells to supply their homes with water.

STUDY OBJECTIVES

  The major objectives for this investigation were the following:
•To establish the source(s) and aerial extent of contamination
•To identify the contaminants and their concentrations throughout
 the study area
•To determine contaminant migration pathways within the bed-
 rock
•To assess the hydraulic conductivity of the bedrock aquifer
  These objectives were attained by the completion of a field in-
vestigation which included an air photo analysis, a fracture trace
analysis, a seismic refraction survey, the installation of monitoring
wells and  a pump test of the water supply well serving the apart-
ment complex. The results of this investigation are discussed in this
paper.

GEOLOGIC SETTING
Bedrock Geology

  There have been several studies of the bedrock of the Manchester
quadrangle. The rocks of the area were named, dated and described
on the State of New Hampshire Bedrock Geology Map1 and the
detailed bedrock map of the Manchester quadrangle.1 A bedrock
fracture trace analysis was conducted in Londonderry in 1981. The
most recent studies are being contracted by the Nuclear Regulatory
Commission.'
  The regional trend of the bedrock is northeast. Much of the rock
in the area has been intensely folded with northeast southwest tren-
ding fold  axes. The site lies on the southeastern  flank of the Mer-
rimack synclinorium.
  Billings' mapped the bedrock in the study area as part of the
Merrimack group of meta-sedimentary rocks. Sriramadas1 broke
the Merrimack group into two major subdivisions; the Berwick and
Elliot formations. The Berwick formation underlies the study area.
  The stratigraphic units present at the site can be traced con-
tinuously  from  eastern Connecticut to southern Maine4 (Fig. 2).

-------
                                    EXPLANATION

                             lurian        Pre-Ordovician

                               Etiot  fm.   msl undifferentiated
                               Kittery fm.    metasediments
                               Rye fm.     |bb| Bigelow Br. fm
                                             Paxton Group
                                             Oakdale fm.
                                        Dundifferentiated
                                        int rus i ves
                                     f fault
                                     f contact


                          Figure 2
    Map Showing Generalized Geologic Structure of Southeastern
      New Hampshire and Adjacent Maine and Massachusetts
               (after Barosh and Moore, in prep.)
Barosh, in recent detailed studies, has named and dated these for-
mations. The lower member of the Berwick formation corresponds
to the Paxton group. Recent age dating places these rocks as pre-
Ordivician. In 1953 and  1954, Billings and Thompson noted the
similarities between the lower  member of the  Berwick formation
and the Oakdale formation seen in Central Massachusetts.
  A  major structural feature in the area is the Nashua trough, a
zone of numerous faults of the Paleozoic  epoch.4 One of  these
faults that  appears to  run through the  study area is called the
Nashua fault (Fig. 3). The  fault in this area is concealed by  thick
glacial deposits. To the northwest of the fault are the older, highly
metamorphosed strata of the Oakdale formation which border the
younger, low to  moderately metamorphosed strata of the  Paxton
group.3  It is a northeast  trending fault which dips steeply to the
northwest.  It is  described  as a high  angle,  right lateral,  reverse
fault.
  The Oakdale formation consists of metasiltstones, phyllites and
sillimanite becoming garnet schists. The Paxton group is composed
of chlorite-grade, gray metasiltstones, phyllites and minor amounts
of calcareous metasiltstones.5 Numerous intrusions of quartz mon-
zonite,  granite,  granodiorite  and pegmatites have  been found
throughout the area.3
  The integrity of the bedrock beneath  the site was investigated
with borings and by a seismic refraction survey. The bedrock was
highly weathered and fractured throughout most of the study area.
The bedrock has been affected by both the large scale geologic pro-
cesses, which caused the metamorphism and deformation of the
rocks, and smaller scale occurrences such as faulting. In addition,
the Berwick formation is brittle in nature and glaciation has added
to the fracturing and caused substantial weathering of the bedrock.
  A  study of the bedrock fracturing just north of the study area
shows predominant northeast and northwest fracture orientations.
At an outcrop in a road cut one mile north of the site, the majority
of fractures had a strike of N45E and a dip of 60° NW. In a frac-
ture study by Ecology and Environment, Inc., geologists produced
a stereographic plot of the poles of 165 fractures. The majority of
the fractures had a N57E orientation.
  The elevation of the top of bedrock  above MSL is  shown on
Figure 4. The contours which are not in the boldprint are land sur-
face  elevations above MSL (Refer to Fig. 3). The general slope of
the rock is in a southeasterly direction with a 4% slope.
  If the surficial deposits on the site are heterogenous,  hydraulic
conductivities vary from place to place. Although the sediments on
this site are primarily glacial stream and alluvial deposits, they con-
tain a mixture of grain sizes. Sizes range from coarse to fine sands
with traces of gravel and silt. In addition,  the till contains cobble
size materials and clay.
Groundwater Hydrogeology

  Unfractured  schist has a low hydraulic conductivity (1 X 10 ~6
cm/sec  and less).  Schistose rocks contain  planes  of  weakness
paralleling foliation which promotes fracturing and erosion. Thus
zones of higher conductivity often lie parallel to the foliation. The
foliation trends northeast and dips steeply  to the northwest in the
site  area.  Fractures, joints and  other discontinuities provide
pathways for ground water flow. This became apparent during the
drilling.  The unconsolidated aquifer was found  to be in good
hydraulic communication with the  bedrock  aquifer. The rock is
highly weathered in some areas. This leaves an extensive void space
for the ground water to fill.
  The predicted groundwater flow direction in the unconsolidated
aquifer is to the south and southeast. The flow direction is strongly
influenced by the  bedrock topography of the site. The recharge
areas lie to the north of the site. The major discharge zones are a
tributary to the Merrimack River and the swamps bordering it.
Water Levels and Gradients

  Data  from the 31 monitoring wells are presented in Table 1.
These data include total depth, depth to bedrock and water level
elevations. Cluster wells have the deeper well designated by the well
number, e.g., wells 10 to 10D.
  A contour map  of the water table elevations  in the shallow
bedrock wells is presented in Figure 5.
  The flow gradients are southeast  in the  area behind the garage
and south for the remainder of the site. The upper 15 ft of bedrock
are fractured and weathered over a large portion of the study area.
This  condition allows  good hydraulic  communication with  the
overlying glacial deposits. Therefore, using equal head contour
lines, flow direction can be approximated  in the shallow bedrock
aquifer. However, this  is not possible in the deep, bedrock wells.

FRACTURE HYDROGEOLOGY

   Many papers have been written about the study of fluid move-
ment through porous media. This  is not the case  with fracture
hydrogeology. Little attention has been given to this field because
of the lack of economic incentive and the extreme complexity of the
problem.
   One of the first major experiments in fracture hydrogeology was
conducted on  artificial fractures.7  A comprehensive  review of
literature in this field' indicates there is little agreement on the ac-
curacy of the existing models for flow in fractured media.
   The majority of crystalline and sedimentary rock masses consist
of rock blocks bounded by discrete fracture planes.' The study area
consists  of a  metasedimentary  schist with  low  porosity  and

   CONTAMINATED GROUNDWATER CONTROL      151

-------
                                                                                                            Key
                                                                                             !{?*•
                              vv
                                                                                         i Quartz MonzanKa
                                                                                         k, A mMvnorpliOMd
                                                                                                     Paxton Group
                                                                                                     Oakdate Formation
                                                                      Od
                                                            Figure 3
                                                          Bedrock Map
permeability. Since the hydraulic conductivity is estimated to be be-
tween 1 x 10-2 and 1  X  10 ~7 (cm/sec), flow  through the rock
itself would be insignificant. Thus, the major flow must occur via
secondary porosity. The pathways for groundwater flow and con-
taminant migration in bedrock to be considered are joints, fracture
zones and shear zones.  The flow through these discontinuities is
controlled by the orientation, interconnection and spacing of these
linear features. To adequately address the directional permeability
of the rock mass, all of the  above characteristics must be con-
sidered.
  Joints can be  considered effective  pathways if their ratio of
length to joint spacing  is large. If they are interconnected,  they
form  a potential flow network. The extent of interconnection is
dependent upon  the orientation of and  length of the different joint
sets.
  Fracture zones may consist of closely spaced, highly  intercon-
nected fractures  which have remained  open. These fracture zones
are an aggregate of fracture sets. A fracture set can be defined as a
number of fractures having the same or closely shared orientations.
Fracture sets can be identified by plotting  the poles of  the in-
dividual fracture planes  on a steronet.  The resulting plot indicates
the orientations  of the various fracture sets. Recent studies have
found that fracture orientations measured on the surface (i.e., out-
crops) have similar  orientations to  those in the  subsurface.10'"
Permeability is also determined by the degree of interconnection of
fractures and the distribution of the fracture aperatures.
  The principles of flow and permeability of fractures are of great
importance. The primary fracture characteristic is the size of the
aperature which exerts a major influence on flow. The aperature of
the fracture cubed is proportional to the flow through a unit length
                                                                                                  Elevations of surface
                                                                                                  Top of bedrock above MSL
                                                                                    Figure 4
                                                                               Bedrock Elevations
152
CONTAMINATED GROUNDWATER CONTROL

-------
                                                                     tion.'" As stress increases, hydraulic conductivity decreases. It is
                                                                     thought that interconnections and fracture aperature decrease with
                                                                     increases in stress which, in turn, reduces hydraulic conductivity.

                                                                     Pump Test
                                                                       Evidence supporting the theory that contaminant migration is
                                                                     fracture  controlled in the study area was  gathered during  the
                                                                     groundwater study. After analyzing the results of the groundwater
                                                                     sampling and mapping the oddly shaped plume of contamination, a
                                                                     pump test was initiated.
                                                                       A 68-hr pumping test was conducted at the site on Aug. 9, 1983,
                                                                     and water  level  elevations were recorded in all the wells for the
                                                                     duration of the test. The apartment complex supply well was the
                                                                     pumping well, and a discharge rate of 20 gal/min was continuous
                                                                     for the entire test.

                                                                     Flow Conditions in Fractured Rocks

                                                                       A pump test in an unconfined homogeneous aquifer causes a
                                                                     symmetrical cone of depression in the surrounding aquifer. The
                                                                     flow lines converge on the well from all directions, and drawdown
                                                                     decreases with distance from the pumped well.
                                                                       The results of a pump test in fractured bedrock are very dif-
                                                                     ferent. When a  well which intersects a fracture  is pumped, the
                                                                     water level in this fracture is drawn down. Continued pumping in-
                                                                     duces  movement of the water into the fracture from other, inter-
                                                                     connected fractures within the aquifer. The pumped fracture acts
                                                                     much  like a collector well. The drawdown is not radial but is a
                                                                     trough-like depression parallel to the pumped fracture (collector
                                                                     well),  as shown  in  Figure  6.  Radial flow equations do  not ap-
                                                                     propriately apply to pumped fracture drawdown data.14
                                                                       When plotting pumping test data where linear flow is suspected,
                                                                     a  semi-log of drawdown(s) versus time(t) is not appropriate. The
                                                                     use of a semi-log plot of drawdown(s) vs. time(t) will result in a cur-
                                                                     vilinear plot. This indicates that the traditional methods of aquifer
                                                                     test analysis are  not applicable since the aquifer is extremely
                               Key
                                              E"qui potential
    RW2   Residential Well

    FW3   Reid Investigation Team (Shallow Well)*

    FW4   Field Investigation Team (Deep Well}"

    ERT5   Emergency Response Team Deep Well

    LGSW  Londondeny Green Supply Well

    LG/WV  Londonderry Green Abandoned Well

    LGEW  Londonderry Green Emergency Well
         Water level elevations above MSL in shallow
         bedrock—data from wells less than 50 leet into
         bedrock
•Shallow Wells  Consist of an open borehole in
         bedrock up to 50 feet Into rack

 • • Deep Wells  Consist of an open borehole in
         bedrock from 100 feet to 300 feet into
         rock

^^^^^^  Direction of flow potential
                           Figure 5
                   Water Level Elevation Map
of the fracture. This relationship explains how a few major frac-
tures can dominate the entire flow system.
  Fracture intensity (degree of fracturing) has a strong effect on
the directional permeability of the rock mass. The orientation and
spacing of fracture sets  define the  fracture intensity. A similar
orientation among a major portion of the fracture sets will show an
increased permeability in that direction. Directional permeability is
not geometrically related to lithologic boundaries in a frequently
anisotropic fractured rock flow network.12 Shear has been studied
the least although it has a  great effect on permeability.13  The
hydraulic characteristics of fractured rock masses and the relation
to stress and depth should be considered. As depth  and stress in-
crease,  hydraulic  conductivity decreases in  a  linear  fashion.
"Although it is assumed that stress  and  fracture  interconnection
may be the two most important variables controlling flow through
fractured rocks, no quantitative data exist to verify this  assump-
1
1

1
1
1
I
1
r -j~
i
i
i
i
i
i
/
•*•
2
u.

_
1
J
i
I
1
1
1
i
1
•
5 i
* i
1
1
Pumped v
j
-o f
-5 1
S '
Ł '
i
1
i
,
^

i j
r~ -J
•'el,
'



                                                              Figure 6
                                              Conceptual Model of a Linear Flow System14
                                                                       CONTAMINATED GROUNDWATER CONTROL
                                                                                              153

-------
                                                              Table 1
                                                Water Level Reading) During Pump Tent
                                                (Measured in feet to top of casing, TOC)
Well
Numbers

FW-OI
FW-02
FW-02D
FW-03
FW-03D
FW-04
F»-03
FW-06
F»-07
FW-OS
FW-08D
FW-09
F»-IO
FW-IOD
FW-II
Fw-lID
F»-I2
F*-I3
FW-14
FW-13
FW-16
FW-17
FW-ll
FW-19
FW-20
ERT-OI
ERT-02
ERT-03
ERT-04
ERT-06
ERT-OI
LCSW
LCEW
LCAW
ELAPSED TIME
O Hours
TOC
9.9
6.1
7.4
8.4
10.1
7.3
6.4
10.7
3.5
11.4
11.2
10.3
t.l
24.6
6.9
14.6
II.)
10.7
1.1
15.1
1.1
3.9
7.1
3.0
6.1
2.7
6.3
2.1
17.9
7.1
1.7
2.1
Artesian
Artesian
Elevation
317.63
211.77
219.78
283.49
211.03
275.86
210.0)
260.21
238.51
262.1)
262.9)
296.74
282.29
267.76
273.67
268.22
268.0)
268.78
264.20
269.03
246.69
229.34
246.31
244.06
275.00
281.64
278.04
254.55
248.84
247.36
298.08
266.57
>266.«3
>246.73
24 Hours
TOC
NM
6.3
7.8
8.1
ll.l
10.8
7.6
ll.l
3.3
14.6
)).4
10.6
9.1
50.2
7.2
40.0
D.I
11.0
16.9
13.2
8.3
4.0
7.2
3.0
6.7
3.8
9.5
3.0
18.1
7.2
9.1
54.6
»6.2
Artesian
Elevation
NM
288.37
289.38
28). 09
279.3)
280.33
278.83
239.81
258.31
258.9)
240.7)
296.44
281.99
242.16
275.37
242.82
266.2)
268.41
255.40
268.93
246.49
229.24
246.41
244.06
274.40
280.54
274.84
254.33
2*1.64
247.26
297.61
214.07
220.2)
>246.73
41 Hours
TOC
10.1
6.)
7.7
1.6
11.3
10.4
7.6
11.5
). 3
15.4
37.)
10.6
9.)
36.7
7.4
42 t
14.2
II. 0
19.6
13.4
1.2
3.9
7.2
4.9
6.9
3.)
9.2
3.6
ll.l
7.3
1.1
57.0
•1.7
Artesian
Elevation
317.43
281.37
219.41
213.29
279.65
210.75
271.1)
259.41
231.31
231.1)
236.13
296.44
211.79
235.66
273.17
2*0.22
265.13
261.48
232.70
261.73
246.39
229.3*
246.41
244.16
274.20
211.04
273.14
233.73
241.6*
247.16
297.91
211.67
217.73
>2»6.73
61 Hours
TOC
NM
6.)
7.1
I.I
12.0
10.1
7.)
11.7
1.6
13.3
37.9
10.7
9.3
J7.I
7.)
«).6
14.3
tl.O
20.0
13.3
1.)
3.9
7.2
4.6
7.0
3.5
9.6
3.7
ll.l
7.2
I.I
37.6
49.*
Artesian
Elevation
NM
218.57
219.31
283.99
279.1)
210.) 5
279.1)
239.21
231.41
251.0)
236.2)
296.34
281.79
2 3*. 36
273.27
239.22
264.13
263.48
232.30
261.1)
2*6.49
229.3*
2*6.41
2**.«6
274 10
283.14
27t 74
233.63
248.6*
2*7.26
297.98
211.07
217.03
>2*6.73
Total
Drawdown

0.2
0.2
0.4
0.4
1.9
3.3
0.9
1.0
O.I
4.1
26.7
0.4
0.3
33.2
0.4
29.0
3.2
0.3
11.9
0.2
0.2
0.0
O.I
-0.*
0.9
O.I
3.3
0.9
0.2
O.I
O.I
33.5

0.0
                   Pumping Well:  LOSW
                   Suiting Time  1600 hours
                   Starting Dale: August 9, 1983
                   Discharge, Gallons per Minute (GPM): 20
                                           NOTES:
                                            NM - Not Measured
                                            Rain began at approximately 0900 hours on Aug. II, 1983 Ml hours elapwd lime) and continued
                                              through the end of the pump lest. Approximately 1.25 inches of precipitation fell during this
                                              time which may have casued WIICT levels to rise in some wells, particularly in shallow wells.
                                              during this lime period.
                                            See Figure G-2 for locations
                                            From draft NUS Report, 16
 anisostropic. A plot  of drawndown(s) vs. time(t) on arithmetic
 graph paper will result in a straight line indicating that drawdown
 vs. time is a linear relationship. Drawdown occurs as a "trough" of
 depression, suggesting linear flow in fractured rock."
   Drawdown was greatest in wells closest to the fracture (collector
 well) and was not dependent upon the distance from  the pumping
 well. That is, the well closest to the pumped well may not be the
 well closest to the fracture collector well.
   Drawdown data from two observation wells on the same side of a
 fracture can be used to determine two possible orientations of the
 fracture" (Fig. 7). Field data can then be used to support the most
 probable orientation.
       Pimped
                                          Fracture
        well
                           Figure 7
   Comparison of Radius (r) from a Pumping Well and Distance (x)
          from an Extended Well in an Idealized Aquifer."
                                                             The equation developed" was used to calculate drawdown at any
                                                           perpendicular distance, x, from the fracture at any time, t, after
                                                           pumping. A first order McLaurin expansion performed to simplify
                                                           the results and adapt them to an arithmetic plot of drawdown (s)
                                                           vs. time(t) yields the equation:"

                                                                                                                       (1)
                                                             In this equation: t,, =  time 0, the value of (2 vT> or t at the line
                                                           of zero drawdown; S = storage coefficient; T = the transmissivity;
                                                           X = the perpendicular distance in feet from the fracture of extend-
                                                           ed well from equation (1).
                                                             The plotted data of each observation well is displaced from the
                                                           origin  along  the  v/T  axis  according  to  the  hydraulic
                                                           diffusivity of the system and the perpendicular distance from the
                                                           pumped  well.
                                                             The radii  from the pumping well and the angle between the two
                                                           observation   wells can  be  measured  in the field  (Fig. 7). The
                                                           unknowns are the two perpendicular distances (XI, X2) from the
                                                           fracture and the angles between the radii (Rl, R2) and  the fracture.
                                                           These values can  be determined using the following equations:
                                                                 XI =  Rlsinflfl
                                                                 X 2 = R2 sin 02
                                                                 Ajj! = 01 -  02
(2)
(3)
(4)
                                                             As discussed earlier, drawdown is proportional to the distance
                                                           from the fracture surface. Therefore:
154
CONTAMINATED GROUNDWATER CONTROL

-------
  and
t


XI


X2

XI
               02

              X2
                R2 Sin 02

                Rl Sin 01
Solving for 02: 02 = Tan-1
  and 01   02 +
                            (7)
——°Ł_ )

R2  fcol  - M '02  ^ &
                                                    (5)
                          (6)
                                                          (7)
Then XI and X2 are obtained from equation (2, 3 and 4),  and
Equation 7 becomes:
                Rl t  ,   Sin /\0
    02   tan-1 (	5?	•=_ ,
                Rlv to2  Cos A0-R
                                                          (8)
  Data from wells 8D and 10D were used in the above equations.
The observation wells appeared to be lying on the same side of the
collector  well  and  had  similar slopes with  good  drawdown
response.  The  t0 intercept for well 8D is  36 min, and  the ex-
trapolated t0 for well  10D is approximately 1 min. From field
measurements, the AO is 39 ° and the radii from the pumped well are
300 ft for  8D and 1,250 ft for 10D.
From equation (6)

From equation (7)
                   't  10D
                  	°	  =	0.166
                   t  BD       36
                 300 feet x 1/1 min. x Sin 39°
0 10D = tan-1 [	
              1250 feet x ^36 min. - 300 feet 1 min. cos 39

ThenOSD =  39° + 1.5° =  40.5°
From equation (2) XI =  300 ft. sin 40.5 =  194.9 ft
From equation (3) X2 =  1250 sin 1.5 =  32.7 ft

  The probable orientation of the fracture on the map results in a
N50E trending fracture or extended well.  The schist is relatively im-
permeable and can only allow substantial  flows  through  major
discontinuities which are interconnected. Thus the  drawdown is
more a function of how close the observation well is to the fractures
and other discontinuities and not the distance from the pumping
well. For example, well 10D, which is 1,250 ft from the pumping
well, has a drawdown  of 33.2 ft,  while well  ERT 2, at a  500 ft
distance from the pumping well, has a drawdown of only 3.3 ft.
  After plotting  the zone of increased drawdown (  25  ft) as
shown on Figure 8, a trough of depression  becomes evident. The
trough or zone of depression trends approximately N50E from the
pumping well. The observation wells with the deepest drawdowns
are closer to the fractures and their interconnections.
  The observation wells most affected by the drawdown were the
deep wells that  are 50 to 100  ft into bedrock.  The deepest
drawdown was seen in two of the deep wells, 10D and 11D, which
are located in the source area south of the garage.
  Although the flow to the pumping well is linear in the local area,
the regional flow is a radial flow. The fracture or collector well is a
finite element in an aquifer that can be considered infinite.
PATHWAYS OF MIGRATION
  An understanding of  the pathways  through  which the con-
taminants migrate is based on a knowledge of the groundwater flow
regime and the physical characteristics of the contaminated aquifer.
The pathways through which the contaminants are migrating can
be assessed. The surficial deposits in the two source areas are thin
with moderate  permeability. Volatile organic compounds which
were dumped either ran  off into adjacent streams and swamps or
infiltrated rapidly  through the overburden into the underlying
bedrock.
  The bedrock  underlying the  source area south of the garage is
highly fractured and weathered.  This allows for good hydraulic
communication with the  groundwater  in the overburden. The
bedrock below the suspected source area within the apartment com-
plex is not highly weathered and is less fractured. Contamination is
still predominant in the bedrock aquifer at this location. The sur-
ficial deposits around both sources have low to moderate levels of
volatile  organic compounds. The levels of contamination in the
bedrock aquifer beneath the sources are high.
  The groundwater from the site, including  the contaminated
zones, discharges to a tributary leading to the Merrimack and the
surrounding swamps. The potential flow gradients in the shallow
bedrock aquifer would be perpendicular to the equal head contour
lines shown on Figure 5.  A groundwater divide and a local recharge
zone are present in the area east of the garage. The groundwater
flow potential in the garage area is to the southeast, whereas the
flow potential for the rest of the site is southerly. This holds true on
the regional scale,  but on  the local scale the groundwater flow is
controlled by fractures.
  The major fracture sets in the area trend N57E  and are nearly
vertical.  Data  from the pumping  tests show drawdown to  be
greatest in the source area south of the garage. A trough of depres-
sion trends approximately N50E from the supply well to the source
area in  the  vicinity of the garage. The contaminants clearly
migrated in a southeast direction from the dumping zones near the
garage through the fractured bedrock into the drinking water wells.
Further support for this migration pathway is shown by the follow-
ing: (1) the orientation of the Nashua fault to the south of the site is
approximately N50E; this  is suspected to be a probable cause of
sympathetic fractures within the brittle Oakdale formation of the
site area, and (2) the calculations used to estimate a probable  frac-
ture orientation using pump test data,14 suggested a fracture orien-
tation of N50E; this evidence substantiates  the pathway of migra-
tion from the source area behind the garage to the wells along
apartment complex supply well.
CONCLUSIONS
  It is clear  that migration pathways in the bedrock can be very
complex, and an entire study  could be devoted to  this subject
alone. Tracer studies, or vertical seismic profiling,  along with ex-
tensive pumping tests would help decipher the intricate intercon-
nected flow patterns at this site. This was beyond the scope of this
study, which  was to establish a hydraulic connection between the
contaminated apartment complex supply well and the source of
contamination.
  The pathways of migration and zone of contamination trend in a
northeasterly direction from the apartment complex supply well to
the source area behind the garage (Fig. 9). The indicators for this
route of migration in the bedrock are the  following: (1) fracture
trace studies in  the nearby area show the majority  of fractures to
trend N57E,  (2) the pumping test data, when plotted, showed a
linear drawdown with the greatest  drawdown in the source area
1,500 ft  away;  the linear drawdown and trough of depression is
representative of flow in fractured  rocks (the  trend here was ap-
proximately  N50E) and (3) the Nashua fault which runs to the
south of the study area trends in a northeast direction which would
probably produce sympathetic fracturing in the brittle quartz abun-
dant Oakdale formation.
  The local flow in bedrock is controlled by the pumping of the
apartment complex supply well. During nonpumping conditions,
the flow will return to a south and southeast direction similar to the
regional flow of the drainage basin.
  The following studies  should be completed to further assess the
complicated hydrogeology of this site:
                                                                   CONTAMINATED GROUNDWATER CONTROL
                                                                                                                         155

-------

                                                             «•*
           •*»
                                    K«y
e
         0-2

        FW100
         333

        ERT01



        LOSW
         M5


        LQEW
         48 <


        LQ»V
         OX)
                                       dmwtownmfMl
                              OMp b*dnx* wfd. Londondany QrMn Supply VM
                                      DMp DMrod. Ml.
                                  Umdondtny Qntn Eimrgtney V
                                      OMp todnxk M««,
                                  Undond»ty Omn AB«»to>Mwn
                              Zon* wim dnwdown ^MUr thanor cqutf u> 25 (Mt
                                Figure 8
                        Drawdown During Pumping
                                                                                        Itey
                                                                                           ,  11
                                                                                      Figure 9
                                                                                 Contaminant Plume
     •The rates and local flow patterns of the contamination must be
      defined through conducting on-site or falling head permeability
      tests on all surficial and bedrock units throughout the site.
     •Radionuclide tracing  or brine slug induction tracing methods
      should be conducted  to provide orientation, spacing  and inter-
      connectedness of the fracture sets in bedrock.
     •Caliper logging and vertical seismic profiling could be used to pro-
      vide or check on the above information as well as to indicate the
      various depths of fractures in the aquifer.

     REFERENCES
      1. Billings, M.P.,  Geology of New  Hampshire Part II—Bedrock
         Geology,   New  Hampshire  Planning  and  Development Com-
         mission, 1956, 203.
                                                           2. Sriramadas,  A.,  The  Geology of  the Manchester  Quadrangle,
                                                             New  Hampshire-Bulletin  No.  2,  The New  Hampshire Depart-
                                                             ment  of Resources  and   Economic Development,  Concord,
                                                             NH, 1966.
                                                           3. Barosh,  P.J.  and  Smith,   P.V.,  eds..  New  England Seismo-
                                                             tectonic  Study  Activities during fiscal year 1981: U.S. Nuclear
                                                             Regulatory  Commission Report,  NUREG/CR    3253,   1983,
                                                             116-123.
                                                           4. Barosh,  P.J.,  New  England Seismotectonic  Study  Activities
                                                             during fiscal year  1977:  U.S. Nuclear  Regulatory  Commission
                                                             Report, NUREG/CR - 0081, R6A, 1978, 138.
                                                           5. Barosh,  P.J.,  Fahey,  R.J.  and Pease, M.H., Jr.,  The Bedrock
                                                             Geology  of the Land Area of the  Boston 2  Sheet, Massachu-
                                                             setts,  Connecticut,  Rhode   Island,  and  New  Hampshire: U.S.
                                                             Geological Survey, open file report 77-28S,  1977.
      156
CONTAMINATED GROUNDWATER CONTROL

-------
 6. Pease, M.H., Jr.  and Barosh, P.J.,  "Distribution and Structural
   Significance of the Oakdale  Formation in Northeastern Connecti-
   cut," in Boothroyd, Jon. C., and Hermes, O.D., eds., Guidebook to
   Geologic Field Studies Rhode Island and Adjacent Areas: 73rd Annual
   Meeting,  New  England  Intercollegiate  Geological  Conference,
   Kingston,  RI,  1981.
 7. Lomize,  G.M.,  Fluid Flow in Fractured Rocks:  Goseneroizdat,
   Moscow, 1951.
 8. Witherspoon,  P.A.  and Gale, J.E.,  "Mechanical  and Hydraulic
   Properties of Rocks Related to Induced Seismicity," Eng. Geol,  11,
   1877, 23-25.
 9. Gale,  J.E., Geological  Society of America:  Special Paper 189,
   1982, 163-181.
10. Kendorski,  F.S.  and Mahtab, M.,  "Fracture Patterns  and Ani-
   sotropy of  the San Manuel Quartz Monzanite," Assoc.  of Eng.
   Geologists J.,  13, 1976, 23-52.
11. Raven, K.G. and  Gale, J.E., Project 740057;  Subsurface  Contain-
   ment of Radioactive Waste: Ottawa, Geological Survey of Canada,
   EMR-GSC-RW Internal Report, 1977,  1-77.
12.  Papadopulous, I.S.,  "Nonsteady Flow to a Well in an Infinite Ani-
    sotropic Aquifer," Proc., 24th Geological Congress: Ottawa, Canada,
    1967, 89-99.
13.  Maini,  Y.N.T., On-Site  Hydraulic  Parameters  in Jointed Rock—
    Their Measurement and Interpretation,  Ph.D. dissertation, Imperial
    College, University of London, 1971, 312.
14.  Jenkins, D.N. and Prentice, J.K., "Theory for Aquifer Test Analy-
    sis in Fractured Rocks under Linear (Non-Radial) Flow Conditions,"
    Groundwater, 20, 1982, 12-21.
15.  Carslaw,  A.S.  and  Jaeger, J.C.,  Conduction of Heat  in Solids,
    Oxford University Press, 1959, 510.
16.  Draft Field Investigation Report, NUS  Corporation, Bedford, MA,
    1984.
17.  McGlew,  P.J., Hydrogeologic Investigation of a Hazardous  Waste
    Site in  Southern New Hampshire,  Graduate Thesis,  Boston Uni-
    versity, 1984.
18.  Remedial  Action Master Plan,  prepared  for the USEPA by NUS
    Corporation, Bedford, MA, 1983.
                                                                         CONTAMINATED GROUNDWATER CONTROL       157

-------
                NDT LOCATION OF CONTAINERS BURIED
                        IN  SALINE  CONTAMINATED  SOILS

                                       ROBERT M. KOERNER, Ph.D.
                                       ARTHUR  E. LORD,  JR., Ph.D.
                                                Drexel University
                                           Philadelphia, Pennsylvania
INTRODUCTION
  The investigation of sub-surface objects can be approached in
two very different ways. The first approach utilizes a destructive
test method such as: test pits,  excavation trenches, auger holes,
core borings and observation wells. While one does indeed "see"
the subsurface materials as they are excavated for ease of examina-
tion   or  subsequent  testing,   such  methods are  not  without
drawbacks  in identifying and  locating buried containers. Some
disadvantages of destructive test methods for this purpose are:
•The information obtained is discontinuous over the area investi-
 gated
•Permission to  enter the properties in question (and  excavate
 therein) may be troublesome or impossible to obtain
•Access for excavation equipment may not be possible at the site
 in question
•Costs are generally high, e.g., small amounts of excavation can
 easily be $300/yd3, and boring costs of $15/ft are not uncommon
•There is a danger to personnel  and to the environment due to
 such materials emptying out of the containers if they have been
 ruptured or pierced
  The second approach used in locating buried containers is the use
of a suitable non-destructive testing (NDT) method(s). Within this
category are the following methods which have been used, or seem
to have general applicability: seismic reflection, seismic refraction,
electrical resistivity, electromagnetic (conductivity) induction, in-
duced polarization, eddy current (metal detector), magnetometer,
continuous  microwave (CW),  pulsed  radio  frequency  (ground
penetrating  radar), infrared radiation  and  sonar  (pulse  echo
acoustics). All of the above methods are not equally suited  for iden-
tifying and locating buried containers, but many are, and the in-
terest in these NDT approaches to the problem of subsurface in-
vestigation seems to be increasing.
  Based on  the authors' past work (which will  be described later),
four of these NDT techniques have direct applicability in the detec-
tion and location of buried containers. These techniques are: metal
detector  (MD),  electromagnetic  induction (EMI),  ground
penetrating radar (GPR) and magnetometer (MAG). Each method
will be described briefly.
  The MD and EMI methods are both inductive methods. A trans-
mitting coil sends a continuous electromagnetic  signal to a receiving
coil.  The signal arrives at the receiver through two major paths.
One path is through the air and does not change with the search
position. The other path is through the subsurface material and is
affected mainly by the local electrical conductivity of the subsur-
face media. If an anomaly in  the subsurface  conductivity is en-
countered, e.g., a buried metal  drum, the signal received through
                                                    the earth path is changed significantly and the instrument indicates
                                                    accordingly.  These methods  are  described in more detail  in
                                                    references 1 and 2.
                                                      The GPR method operates on exactly the same principle as or-
                                                    dinary aircraft radar. A pulse of electromagnetic radiation is beam-
                                                    ed into the ground by a special antenna, and reflections occur from
                                                    any discontinuity  in dielectric constant. The reflected pulse arrives
                                                    back at the receiving antenna and  a display of reflected intensity
                                                    versus depth is presented on an oscilloscope and on a recorder. This
                                                    technique is described  more  fully in Reference 3.
                                                      The MAG method measures the local magnetic field (essentially
                                                    the earth's field) and with it any changes in this magnetic field. The
                                                    type used in this  study is a proton procession model. The local
                                                    magnetic field is determined by measuring the procession frequency
                                                    of the proton magnetic moment. This rate is linear in the magnetic
                                                    field, and as the  frequency can be measured very  precisely, the
                                                    magnetic field can also be measured very accurately.  A steel drum,
                                                    being ferromagnetic,  changes the local value from  the earth's
                                                    magnetic field and, hence, can be detected. The MAG technique is
                                                    described in more detail in Reference 4.
                                                    PRIOR STUDIES

                                                      While each of these methods (and others for that matter) will
                                                    work under ideal conditions, the typical site where waste material
                                                    containers are buried is far from ideal. Rather than burial in dry
                                                    granular soils, drums are usually dumped in  swamps, mudflats,
                                                    water and the like. Furthermore, most of the  successful methods
                                                    the authors have worked with are based on electromagnetic prin-
                                                    ciples, thus soil/water conductivity is a major issue. High conduc-
                                                    tivity areas, e.g.,  storage areas, junk yards or ocean water, can
                                                    severely  influence the techniques. To what extent,  however, is
                                                    largely unknown.
                                                      With these thoughts in mind, a series of test sites were obtained,
                                                    containers of various sizes, burial depths, geometric arrangement,
                                                    etc. were carefully  placed, backfilled and then located with the
                                                    various NDT methods.
                                                      The first site was in a nearly ideal dry sandy soil in an open and
                                                    isolated  field.1  This site provided an  excellent starting point and
                                                    essentially narrowed the authors' thinking from the many NDT
                                                    methods available to the specific four mentioned previously. Steel
                                                    containers buried to 10 ft depths were accurately located; they
                                                    could possibly have been located deeper if stable burial pits could
                                                    have been excavated. Some plastic containers were also located, but
                                                    with poorer results. Various container arrays and the boundaries of
                                                    a "trash dump" were accurately located.
158
CONTAMINATED GROUNDWATER CONTROL

-------
  The second site was much more formidable.' Here a saturated,
silty clay  soil overlying shallow shale rock was  used. Detection
depths were  much  shallower, approximately 4 ft,  and the tech-
niques were troubled by the large amount of background metal in
the area (e.g., trailers, equipment, fences, etc.). Results of the four
methods were reasonable within these limitations  and restrictions.
  Recognizing that containers are sometimes dumped directly into
water7 and that  the salinity of the water can range from fresh to
brine, the third study was directed at these conditions.1 Containers
were  placed  in water on the bottom soils at  four  different sites
where the salinity of the water ranged from fresh to ocean.  To
depths of 3  ft of water  above the containers, the  detection and
delineation results were "excellent" to "no good" in direct propor-
tion to the increase in water salinity.
  Bearing directly  on  these three studies is the  extent to which
groundwater  salinity influences the  detection capability of these
NDT methods.  Questions  about the dominance  of the soil pore
water over  the  solid  particular structure, the  amount  of salt
leaching onto the soil particles as one moves back from the ocean
front, how far from the ocean front these methods can be used,
etc.,  are all unanswered. This study, the fourth in the series, fo-
cuses on the  authors' efforts to answer these questions.
DETAILS OF THIS STUDY

  The site selected  for this study was on an island in a bay at the
southern New Jersey shore. Within a distance of approximately 150
ft from the water's  edge, the ground surface rose to 10 ft above sea
level where it became  relatively level in elevation. The soil was a
medium to fine, granular, sand indigenous to that area of southern
New Jersey. The sand density ranged from loose (near the surface)
to intermediate at a depth of 6 ft, the limit to which we could ex-
cavate.
  All containers were made of steel and varied in size from 1 to 55
gal. They were placed in hand excavated holes and backfilled to ap-
proximately the same density as the adjacent undisturbed soil. The
containers were empty and clean.
  Four separate test patterns were deployed, each with a specific
objective:
•In a low conductivity area, to determine if the results would be
 similar to study  #1  results and  if each  method  was working
 properly. In this regard, two  patterns were used;  one employed
 five 30 gal containers each 25 ft apart at 9, 18, 25, 32 and 50 in.
 cover respectively; the other used containers with a constant cover
 of 32 in (also 25 ft apart) but in sizes ranging from 1,  5, 20, 30 to
 55 gal respectively.
•In a varying  conductivity area  extending inland from the ocean
 front. Here the containers were buried under either 18 or 24 in. of
 cover (at varying distances apart) and varied  in size from 1  to
 30 gal.
•In a high conductivity area to attempt to determine the limits of
 a "trash dump"  measuring 9.3 ft  x 4.2 ft containing a large
 amount of miscellaneous metal items (barrels,  tables, rods, steel
 sections, etc.).
Each test pattern was monitored using the four NDT methods de-
scribed earlier.
RESULTS
  The results obtained with the five different sized containers (55,
30, 20, 5 and 1 gal) under a constant soil cover off 32 in  . are shown
in Figure 2.  On the conductivity  plot  obtained from the EMI
                                                             Figure 1
                                   Photographs of Site Conditions at Barnaget Light, New Jersey and the
                                    Four NDT Methods Used: la. Electromagnetic Induction; Ib. Metal
                                     Detector; Ic. Ground Penetrating Radar; and Id. Magnetometer.
                                                                     CONTAMINATED GROUNDWATER CONTROL      159

-------
if';
            JO     40     «0
     EMI  Mor-oa

"lOO     'JO     '40     itO
                               K>     '00
                                          MO  M«IHO«
'" r ; \ / \
0 JO 40 W
0 W 40 tO
' ' 'i»- ' i":
A
•0
•0
If
A.
OI$'»»CC C'««M
(00 IJO '«0
i»- l»-
                          Figure 2
   Results of EMI, MAG, MD on Varying Sized Containers Buried
                      at 32 in. of Cover
                                       & 901
                                                 I flol
                          Figure 3
 Results of GPR on Varying Sized Containers Buried at 32 in. of Cover

method, it can be seen that the background area is about 15 millim-
ho/m which is similar to that at study site 11. However,  only the
two largest size containers can be detected. Plotted beneath this
response  are the MAG and MD response curves,  both of which
clearly detected all containers except the 1 gal. The GPR trace of
Figure 3 gives a similar response.
  The response curves showing the effect of varying soil cover over
equal  size 30 gal containers are given in Figure 4. Here the EMI and
MD methods are seen to be accurate to about 36 in. depth, while
the MAG method easily detects all containers, even the one buried
at SO in. depth. The GPR trace of Figure 5 accurately indicates con-
tainers buried at all depths.
  In Figure 6, a series of six containers (I,  5, 30,  30, 30, 30 gal)
were buried at varying distances from the ocean itself. The attempt
was to keep all  cover depths at  24 in.,  but the one closest  to the
ocean constantly flooded out, so 18 in.  was used in this case. The
EMI plot shows the conductivity exponentially decreasing from 240
millimho/m at the ocean's edge to  background conditions ISO ft
inland. No containers were located on a plot of this scale; however,
the most  inland 30 gal containers were detected on a plot with a
scale the  same as Figures I and 2. The MAG response  was not
defined until the container at 32 ft was encountered. Further inland
it performed well. The MD was pinned near the ocean front and it
became unpinned only after we were  36 ft inland where it per-
formed well from that point onward. The GPR trace of this scan
(not shown) only picked up the 30 gal container at the 149 ft loca-
tion.
                                                                  Lastly, the "trash dump" of Figure 7 was monitored. Here the
                                                                EMI response shows a high background conductivity at the site of
                                                                30 to 40 millimho/m, but the heavy metal concentration is easily
                                                                noted. Also easily noted were the boundaries using the MAG, MD
                                                                and  GPR (not shown) methods.  While each of the methods in-
                                                                dicated the "trash dump" boundaries, none revealed any detail of
                                                                items within the "trash dump" itself.
                                                                  » f  «
                                                                  M  :
                                                                                                                CMI
                           I!
                                                                                   -l-
                                                                                   40
                                                                                                                M«C  Ml i MX
                                                                                      1C,.     . >0 I-
                                                                                            Figure 4
                                                                      Results of EMI. MAG, MD on 30 gal. Containers Buried at
                                                                                Depths Varying from 9 in. to 50 in.
                                                                                                                   40'
                                                                                            Figure 5
                                                                     Results of GPR on 30 gal. Containers Buried at Depths Varying
                                                                                        from 9 in. to SO in.

                                                                  CONCLUSIONS

                                                                    This study, the fourth in a series, continued the general thrust of
                                                                  burying containers at  known  locations  and depths  and then
                                                                  locating them using NOT methods. Thus ground truth was readily
                                                                  and accurately established.
                                                                    Via the results of the earlier  work, it was seen that  the metal
                                                                  detector (MD), electromagnetic  induction (EMI),  ground pene-
                                                                  trating radar (GPR) and magnetometer (MAG) were the premier
                                                                  NOT  methods to use.  Yet, saturated  clay soils cause even these
                                                                  methods to be difficult to use and interpret. Most important,  the
                                                                  background conductivity levels  cannot be too high with respect
                                                                  to the buried  anomaly  for it to be detected  (exception is MAG
                                                                  method). A study of containers under water in gradually increasing
                                                                  saline conditions clearly showed this to be the case.
                                                                    To  extend this finding, the current study was undertaken where
                                                                  an ocean front condition was used as an electrical conductivity ex-
160
           CONTAMINATED GROUNDWATER CONTROL

-------
                                                              Table 1
                   General Acceptability of Using Various NOT Methods to Locate Buried Steel Containers in Conditions Listed
                                              (Maximum Penetration Depth in  Parentheses)
Soil Type
Granular


Cohesive
Water


Percent
Saturation
0 - 20%
10 - 50%
50 - 100%
50 - 100%
100%
100%
100%
Type of
Void Water
fresh
intermediate
ocean
fresh
fresh
intermediate
ocean
Metal
Detector
excellent (6 ft)
excellent (2 ft)
no good
excellent (4 ft)
excellent (3 ft + )
poor (3 ft)
no good
Electromagnetic
Induction
excellent (10 ft)
average (4 ft)
no good
poor (4 ft)
excellent (3 ft + )
no good
no good
Ground Penetrating
Radar
excellent (10 ft)
excellent (3 ft)
poor (2 ft)
excellent (4 ft + )
excellent (3 ft + )
no good
no good
Magnetometer
excellent (10 ft)
excellent (4 ft + )
-- -
poor (4 ft)
excellent (3 ft*
excellent (3 ft + )
excellent (3 ft + )
treme and conductivity decreased as the survey  moved inland.
Specifically, it was found that:
•For  location  of  individual  buried  containers,  conductivity
 greater than 20 millimho/m seriously impairs use of the electro-
 magnetic based systems, i.e., MD, EMI and GPR
•Also adversely affected in this high conductivity area was  the
 MAG  performance but for obviously different and  unknown
 reasons
•The boundaries of a "trash dump," however,  could not be  de-
  lineated by any of the methods
•At conductivities less than 20 millimho/m, the four NOT methods
 performed well and similar to study site #1  which was  also in a
 granular soil but not in an ocean environment
  As a generalized conclusion, the authors are now in a position to
present a unified view of the use of these four NDT methods to
detect buried containers under a wide variety  of conditions (Table
1) Here it can be seen that conductivity of the background or of the
water in the soil voids  or at  the site itself  presents a formidable
obstacle in use of the methods. Each method performed best in
fresh water conditions (in water itself  or water  in the soil voids) and
performance  gradually  decreased  as   the  salinity  increased.
Therefore, high conductivity is definitely a most  limiting factor in
buried container detection using NDT methods based on electro-
magnetic principles.

ACKNOWLEDGEMENTS
  The  authors express  appreciation to  the  USEPA, Office of
Research and  Development, Municipal  Environmental  Research
Laboratory, USEPA, Edison,  NJ,  for  financial  support under
Cooperative Agreement No. CR80777710. Dr. John E. Brugger is
    2 SO




  I
  V  110


  Ł  IOO
                                            EMI  Method
                  40     60    SO     IOO     120     MO     160     ISO
            2O    40      «     BO     IOO     120     I4O     *0     180
                            Figure 6
    Results of EMI, MAG, MD on Buried Containers as a Function
                     of Varying Conductivity
         f f 40
                                        EMI  M.lhoJ
|| 20

-
-

5


20

I
JO


46

,
40

i

           11300


          [ 55180


          ' MMO-
MAG Mtthftd
           54780
         r '"" r
         I ^ [-
                                            10     ftp
                                        DISTANCE  (fit!)
                           Figure 7
    Results of EMI, MAG, MD in Delineating the Boundaries of a
     "Trash Dump." Scans are Taken Directly Over the Centerline
                     of the "Trash Dump."
the project officer. Graduate research assistants who performed
much of the work were Vincent A. Luciani, George H. Barstar III
and Michael J. Monteleone. Our sincere thanks to all involved.

REFERENCES

1.  Lord, A.E.,  Jr., Koerner, R.M. and Freestone, F.J., "The Identifica-
   tion and Location of Buried Containers Via Non-Destructive Testing
   Methods," Jour. Haz. Mails., 5, 1982, 221-233.
2.  McNeill, J.D., "Electromagnetic Resistivity Mapping of Contaminant
   Plumes," Proc. Management of Uncontrolled Hazardous Waste Sites,
   Nov. 1982, Washington, DC, 1-6.
3.  Bowders, J.J., Jr., Koerner, R.M. and Lord, A.E., Jr., "Buried Con-
   tainer Detection Using Ground Penetrating Radar," Jour. Haz. Mat Is.,
   7, 1982, 1-17.
4.  Tyagi, S., Lord, A.E., Jr. and Koerner, R.M., "Use of a Proton Pre-
   cession Magnetometer to Detect Buried Drums in Sandy Soil," Jour.
   Haz. Mails., 8, 1983, 11-23.
5.  Lord, A.E.,  Jr., et a/., "Use of NDT Methods to Detect and  Locate
   Buried Containers Verified by Ground Truth Measurements," Proc.
   Hazardous Materials Spills Conference, April, 1982, Milwaukee, WI,
   185-191.
6.  Koerner, R.M., et al., "Use of NDT Methods to Detect Buried Con-
   tainers in Saturated  Clayey Silt Soil," Proc. Management of Uncon-
   trolled Hazardous  Waste Sites, Nov. 1982, Washington, DC, 12-16.
7.  	, Ground Penetrating Radar Survey, Elizabeth River, New Jersey,
   Final Report to U.S. Coast Guard, New York, Weston  Consultants,
   June 1981.
8.  Lord, A.E.,  Jr., Koerner, R.M. and Arland, F.J., "The Detection of
   Containers Located  Beneath  Water  Surfaces Using NDT  (Remote)
   Sensing  Techniques," Proc.  Hazardous Wastes  and Environmental
   Emergencies, Mar. 1984, Houston, TX, 392-395.
                                                                      CONTAMINATED GROUNDWATER CONTROL
                                                            161

-------
       CONTAMINANTS  IN GROUNDWATER:  ASSESSMENT
         OF CONTAINMENT AND RESTORATION  OPTIONS

                                                   A.C. BUMB
                                              C.R. McKEE, Ph.D.
                                               J.M. REVERAND
                                          J.C. HALEPASKA,  Ph.D.
                                              J.I.  DREVER, Ph.D.
                                               S.C. WAY, Ph.D.
                                                   In-Situ, Inc.
                                               Laramie, Wyoming
INTRODUCTION

  When contaminants escape from plant  sites due to accidental
spills, leakage from underground storage or from waste dumps, the
potential liability of the site owner is always an issue. If the escap-
ing contaminants enter the ground water, there is great danger that
they will spread and affect the water quality of nearby water-rights
holders. Final liability,  of course, can only be determined in the
courts, when and if the contaminants turn up and the extent of the
resulting problem is determined. This determination may be many
years  in the future because groundwater velocities are generally
very slow. However, there is a quick, easy and inexpensive method
by which a plant owner can determine: (1) whether a contamination
problem exists; (2) when and where it is likely to show up; (3) how
serious its effects may be on the water quality of nearby users; and
(4) how to control and ultimately correct it.
  The owner of a plant that is polluting the air does not have any
difficulty to deciding whether or not a problem exists. He can see
(or measure) the contaminant leaving his smokestacks. But ground-
water is a hidden natural resource and cannot be visually examined
except by drilling wells.  Hence, one often has no clue as to the ex-
istence of a problem. Fortunately, today there  is a very powerful
tool to take the place of visual examination of the groundwater
system under the surface of the earth. Not only does  this tool
enable one to see what would otherwise be invisible, but  it also
enables one  to compress years or decades into minutes or seconds
and predict the future. This very powerful tool is computer model-
ing.

Defining the Problem

  Assessing the potential liability  from accidental contamination
involves answering two questions. First: Is the contaminant finding
its way into the system of underground water? Second: If so, where
is it going? To answer the first question, one must determine the
rate of contaminant seepage from the spill or leak into the ground-
water  system. To answer the second question, one needs to deter-
mine  the velocity and  concentration  of the contaminant as  it
migrates in a "plume"  through the groundwater system. Models
are particularly useful here.
  Contaminants generally  enter the groundwater system  in the
"vadose zone," an area above the water table where the soil is not
saturated with water. With time, rainfall percolating through the
ground leaches the contaminants out of the soil,  and they can reach
the groundwater. There are two ways to determine if this process is
taking place and, if so, in what quantity and at what rate. The first
process utilizes the traditional method of drilling portions of the
                                                    soil and rock and analyzing the drilled cores in a laboratory—a pro-
                                                    cess that can be expensive and time-consuming.
                                                      A quicker, more cost-effective method is to make use of existing
                                                    wells. Pumping several wells takes less time and costs less than drill-
                                                    ing cores.  While the wells are being pumped and  monitored for
                                                    potential contaminants, they are also removing the  contamination
                                                    from  the groundwater (these wells are known as "purge wells").
                                                    These wells  can also be used  later  to monitor the progress of
                                                    cleanup operations, should they be necessary. With projects of this
                                                    type,  the authors have obtained satisfactory agreement between the
                                                    two approaches: core data and purge well data.
                                                      The second option, well pumping, can only be considered if the
                                                    contaminated water can be treated or diluted to meet discharge
                                                    standards. If not, a third method, monitoring concentrations in the
                                                    wells, can  be used.  Using this information together with ground-
                                                    water flow, the rate of leaching can be determined.
                                                      The study of contaminants in groundwater uses geochemistry
                                                    and geohydrology. Geochemical  properties of concern are the  rate
                                                    of leaching of the contaminants to the groundwater system and the
                                                    rate at which the contaminants  are adsorbed or retained by par-
                                                    ticles  of soil. Geohydrologic properties used to construct a satisfac-
                                                    tory model are: the water level and groundwater flow rate; and the
                                                    porosity, transmissivity and density of the soil or rock.

                                                    GEOCHEMISTRY

                                                      Computer modelers describe the rate at  which contaminants are
                                                    released into the groundwater as a "source function." The source
                                                    may be instantaneous, as when all the contamination from an ac-
                                                    cidental spill reaches a shallow aquifer at the same time. The con-
                                                    taminant may enter the groundwater  at a constant rate (for exam-
                                                    ple, from  a slow, steady leak in a pond or  underground pipe or
                                                    storage tank).
                                                      A third  and more complex situation arises when the pollutant
                                                    enters the  unsaturated zone above the water  table and is gradually
                                                    leached out of the  soil into the groundwater system through the
                                                    percolating action of rainfall. The spill  may have been instan-
                                                    taneous, but with time less and less of it is getting into the ground-
                                                    water. This is known as a "decaying source term," because the rate
                                                    at which the contaminant is entering the groundwater is constantly
                                                    decreasing. These source functions are described in Appendix A.
                                                    Adsorption

                                                      A second essential geochemical process is  adsorption. Once the
                                                    contaminant is in the groundwater, one wants to know how fast it
                                                    will move. To visualize the geochemical process of adsorption, im-
                                                    agine two  party-goers walking together down a street toward their
162
CONTAMINATED GROUNDWATER CONTROL

-------
homes at the end. One cannot resist stopping in for a drink at every
bar along the way; his progress toward his home is severely re-
tarded.  The  other— the  "non-adsorbing  species"  in  this
analogy— walks straight  down  the street  and  into his house.
Similarly, some contaminants— nitrate, for example— do not ad-
sorb on clays and sand. Ammonia, however, is strongly adsorbed
by clay minerals, just as the first party-goer was strongly  attracted
to drinking  establishments.
  When  groundwater containing  an adsorbing  species  migrates
through an  aquifer, the species is  first removed from solution by
adsorption on solids,  and then gradually "washed off" as cations
in the fresh  groundwater displace it from solids back into  solution.
As a result, a  plume of an adsorbing species in  groundwater
migrates at a much slower rate than the rate at which groundwater
moves; an effect called "retardation." The magnitude of the retar-
dation effect, R, is given  by
   R    1+M1-*>P9                                        (1)
where  is the porosity,  pg is the grain specific density and kd is the
distribution coefficient  (a ratio of the contaminant concentration
on soil to the contaminant concentration in water at equilibrium).
   The magnitude of the retardation effect depends on how strongly
the species is adsorbed by the sediment, which in turn is a function
of the adsorption properties of sediment and the groundwater com-
position. In general, aquifer materials having a higher clay content
tend to be more adsorbing.
   Important adsorption  parameters  are the distribution  coeffi-
cient, kd, and the cation exchange capacity, CEC (the maximum
amount of species that  can be adsorbed).  The  parameters can be
determined in the laboratory  using a core sample taken  from the
aquifer and natural or synthetic groundwater, or they may be deter-
mined in the field using  a push-pull test.' The  field test is preferable
as it samples approximately 100 tonnes of sediment compared to a
few grams to kilograms in the laboratory.2

Conversion of Contaminants

   Chemical  reactions — e.g.,  oxidation,  reduction,   hydroly-
sys — change,  destroy  or  neutralize  contaminants. Sometimes
the end product  may  pose  a greater  health  hazard  than  the
original  contaminant. Most organic degradation is  carried  out
by microorganisms.  Radioactive materials lose their activity by
disintegration. Many of these conversion  processes can be modeled
using a first-order rate law; i.e., the conversion rate is proportional
to the amount  present. The  model described  in Appendix B is
capable of handling this type of conversion. Conversion of species
is site-specific, and caution should be exercised when applying data
from one site to another.

GEOHYDROLOGY

   Many tests are available to  define the  hydrologic characteristics
of the groundwater system underlying a site.3  Details of the various
procedures are available elsewhere.4'5 Hydrologic properties can be
estimated from pump test data or computer simulation. Hydrologic
properties of interest are the saturated  thickness of the aquifer,
hydraulic gradient, porosity,  hydraulic  conductivity and aquifer
boundaries. These  data can be combined to obtain groundwater
velocity.

Dispersion
   Dispersion, the mixing of miscible fluids as  they flow through
granular media, is another  geohydrologic phenomenon governing
contaminant movement.  Dispersion causes  the contaminants to
contaminate a  larger volume, but at lower concentrations. On a
microscopic scale, mixing of fluids occurs due to molecular diffu-
sion and microscopic or macroscopic variations in flow velocity.
Velocity variations  are due to  heterogeneities which cause the con-
taminant front to spread, both laterally and in the major direction
of flow. In the aquifer  systems of concern, molecular diffusion is
negligible  compared to  the  mixing due to velocity variations
resulting from heterogeneities.
  Dispersive mixing is quantified through the use of the dispersion
coefficient, which is known to be proportional to the groundwater
velocity;' the coefficient of proportionality is known as the disper-
sivity. Dispersion occuring in the direction of mean flow is termed
longitudinal dispersion;  dispersion occurring perpendicular to the
direction of mean flow is termed transverse dispersion.
  Most reported transverse dispersivities are from 20% to 35% of
longitudinal dispersivity. Dispersivity obtained using fracture fre-
quency distributions7 suggests that transverse dispersivity is about
20-40% of longitudinal  dispersivity.  It is  well  known that the
magnitude of measured dispersivity changes depending on the scale
at which the measurements are taken.'
  Geologic heterogeneities also influence the magnitude of disper-
sivity. A greater number of heterogeneities  combined with an in-
creased travel  distance  results  in  larger dispersivity  values.'
Laboratory experiments yield values in the range of 10 ~2 to 1 cm,
while dispersivities of 10 to 100 m  have been obtained for field
problems.6 The larger the dispersivities, the more  the chemical's
concentration will tend to decrease; hence, natural restoration will
proceed at a faster rate. It is believed that natural aquifer dispersivi-
ties increase to a "critical" value,  and then remain nearly constant
as the dimensions of the aquifer increase. Those dispersivity values
obtained from larger-scale history matches of aquifer contamina-
tion are felt to be most applicable to pollution problems.
  Although dispersivities are important, it may often be  imprac-
tical to measure actual dispersivities in a reasonable time at the
necessary length scale. However, it may be possible to use a push-
pull test1 to determine a local dispersivity which can then be scaled
up to the correct length scale using the correlations developed by
Lallemand-Barres and Peaudecerf.10
THE MODEL

  One may choose either a "top-down" or a "bottom-up" ap-
proach to designing a computer model." A "top-down" approach
begins with a great deal of detail and proceeds through stages of in-
creasing simplification. A "bottom-up" approach, as might be ex-
pected, begins with the simplest model and adds details until the
point of diminishing returns is reached. That is, additional com-
plexity does not result in any great pay-off in information gained.
The bottom-up approach is generally faster and nearly always more
cost-effective.
Numerical vs. Analytic Models

  Models may be based on numerical techniques (e.g., finite-
difference, finite-element and boundary-element methods) or on
analytic or approximate analytic solutions. Both methods  have
their drawbacks: numerical methods suffer from instability and are
not practical for the non-specialist; analytic solutions cannot han-
dle  complex heterogeneities in porous  media. Some  numerical
models (such as that of Gupta et  a/.12) do not work when disper-
sivities are less than 80 m.
  Frequently, all the data required for a numerical solution are not
available,  and there may not be time to collect them before the
pollution problem  becomes  serious. In such cases, analytic solu-
tions are more practical. Simple analytic models can also be used to
run what is called a parametric or sensitivity study to determine the
need for detailed data (determined by the relative sensitivity of the
model  output to various input data). A sensitivity study is one of
the   best  applications  for  these models  due to  their  cost-
effectiveness.
  Whether one chooses  a  numerical or  an analytic model, the
underlying idea is what is known as "the conservation of mass." A
mass balance equation for a given volume at a particular location
may be expressed in words as:
 prate of  1       fr
  change of I   -   If
 |_mass    J       [j
rate of trans-1    Prate of trans-"!    ["sources
port of mass I -  I formation   I
in and out  J    j_of mass    J
   fsources"!
*   of
   |_mass  J
The rate of mass transport into  and out of a location  includes
transport by convection (flow of water) or by dispersion and diffu-
sion. A simply analytic model is presented in Appendix B.
                                                                     CONTAMINATED GROUNDWATER CONTROL
                                                                                                                             163

-------
SENSITIVITY ANALYSIS AND
APPLICATION

  As noted earlier,  a "bottom-up" approach is generally  cost-
effective,  and this is the approach  the authors recommend. The
authors also recommend the use of analytical and quasi-analytical
models wherever possible. Such models are particularly suited for a
number of runs required for sensitivity  analysis. Ammonia and
nitrate migration are used here as an example; other contaminants
can be handled in similar fashion.
  As an example, consider the site shown in Figure 1. It includes a
plant, underground storage tanks and above-ground packing and
materials  handling area. Over the years, ammonia and nitrate have
been spilled. Both have leached from the soil and now appear in the
groundwater system, while 200,000  kg of ammonia are still in the
ground. (It is  assumed that  sufficient preventive measures have
been taken so that  further ammonia and nitrate are not being
spilled.) Management is now faced with the question of whether or
not a pollution problem exists. If the answer is yes, then what is the
most cost-effective remedy? The sensitivity analysis can provide a
quick answer to these questions.
  For this example,  the authors assume that groundwater velocity
is 60 m/yr, the aquifer is 15 m thick and porosity is 30% with a
grain density  of 2.6. As the groundwater velocity increases, the
dispersion coefficient, which is the product of the aquifer's disper-
sivity and groundwater velocity, will increase. However, the time of
arrival of maximum concentration at a fixed point  will decrease,
allowing less time for the plume to disperse (although the plume is
dispersing at a higher rate). The result of these competing effects is
that maximum concentration will be the same. By manipulating the
equations given in Appendix B, one  can rigorously prove that
groundwater velocity does not  significantly  influence maximum
concentrations away from the site, but it does govern the plume's
arrival time at any point. Increasing dispersivity, on the other hand,
will lower the concentration when  the peak arrives at  the  same
selected  point  regardless  of the  groundwater   velocity.  A
longitudinal dispersivity of 35 m and a transverse dispersivity of 12
m  are  selected.  Dispersivities are scale-dependent. Small-scale
dispersivities can be easily measured (for example, by using a push-
pull test1), but larger-scale dispersivities are normally estimated due
to time and budget constraints. Therefore, the dispersivity was not
varied in this example, although it can be easily varied in the model.
   If the groundwater is saturated with oxygen,  and  100% of the
oxygen is used to convert ammonia to nitrate, only 2 mg/1 of nitrate
as N would be present. A 100"% conversion efficiency is unlikely.
Freshwater intrusion and infiltration,  and diffusive transport of
oxygen through the  vadose zone to the groundwater, are possible
mechanisms which will help convert ammonia to nitrate. The oxy-
genated near-surface groundwater would then have to mix vertical-
ly in the aquifer to convert ammonia to nitrate. The authors do not
                          -225
                                                                      40,000



50 m (2)
1 u
1 100 m'






| (0, 0)

(3)
^

inn m
t
Ł
«
1
t
E
n
t

centerlme for
maximum
concentration
calculation
J_





               Ground Water Flow
               — - ».
               60 m/year

                           Figure 1
                Site Layout for Model Calculations
                                                                     30,000
                                                                  
-------
                   NOj
                     time lor maximum
                    -concentration to occur
                     in years
                          1000        1500        2000
                        Distance Along Centerline (m)
                                                          25OO
                            Figure 3
      Maximum Nitrate Concentration as a Function of Distance
                 and Nitrate Leach Rate (half-life)
  The configuration of a plume after 25 years for a source half-life
of 10 years is shown in Figure 4. Nitrate still appears at the site, and
the plume widens as distance from the site increases due to disper-
sion. Dispersion dilutes concentrations and tends to keep maximum
concentration at a low level as distance from the site increases (see
Fig. 3).
  The concentrations occurring at various points along the plume
centerline are shown in Figure 5. It takes some time before nitrate is,
seen down-gradient. Once it arrives at a location, its concentration
slowly increases to a maximum and then begins to decrease. As the
distance increases, maximum concentration decreases,  and  it also
takes longer  for the  plume to disappear.
  In Figures  3,4 and 5, the authors assumed that nitrate leached in-
to the groundwater system uniformly over the site. However, this is
rarely the case. Some areas are generally more contaminated than
others. Should one spend a fortune to define  relative rates of con-
tamination within the site? Figure 6  shows that the answer is no if
one is concerned with liabilities at some distance from the site. If
the nitrate concentration near the site is important,  then one might
consider defining  the  source term more  accurately. The relative
amounts of nitrate in various regions  in Figure 1  and the relative
concentrations with  respect to region  1 were used to define non-
uniform distribution of nitrate (Table 1). For example, one initially
might see 21  mg/1 nitrate as N in region 1, 63  mg/1  in region 2 and
12 mg/1 in region 3. Even in cases where the initial concentration is
much higher, the knowledge of its distribution is not important if
the bulk rate of leaching into groundwater is available or can be
estimated from purge  or monitor well data—the exception again
being if the point of concern is near the site.

Remedial Actions
  Drinking water standards for nitrate have been established at 10
mg/1 as N.13  If the source half-life is  2 years, then potential liability
exists within 1500 m down-gradient  of the site boundary (Fig.  3).
Various courses of action may be considered by management at this
point. Three of these are described below.
Natural Restoration
  The natural mechanisms of groundwater flow and dispersion will
eventually lower concentrations to acceptable limits. In many cases,
concentrations  will  be higher near the  site early;  thus,  some
downstream  water  users  may be  adversely affected, requiring
mitigation or relocation of water sources. However, this impact can
be reduced or eliminated by pumping with purge wells until natural
restoration occurs. The financial consequences of  this  option  will
vary widely from one situation to another.
                            Table 1
       Amount of Nitrate in Various Regions for Nonuniform
                     Distribution in Figure 6
Region in
Figure 1
1
2
3

Nitrate in place
(kg)
40,663
31,627
27,710
Total: 100,000
Concentration
relative to region 1
1
3
0.56

                                                                                    60 m/year
                                                                                                                       25 years  .
                                                                                              500        1000
                                                                                                   X(m)
                           Figure 4
      Nitrate Concentration (ppm as N) Contours after 25 Years
                        (ti/2 = 10 years)
               10    15    20    25    30    35    40    45    50
                           Figure 5
       Nitrate Concentrations at Selected Locations vs. Time for
                         t'/2 = 2 years.
    -250   0    250   500   750  10OO   1250  15OO   1750   2000  2250
                       Distance Along Centerline (m)


                           Figure 6
Downstream Concentrations for Uniform and Nonuniform Distribution
      of 100,000 kg of Nitrate Leaching into Groundwater with
                         t'/j  =2 Years
                                                                      CONTAMINATED GROUNDWATER CONTROL       165

-------
Purge Well
  Install a purge well at an optimum location on the site, using a
computer model to ensure the maximum containment of nitrate
within the site.  Treat  or  dilute the  water before discharging.  A
pump with a capacity of 1.1 mVmin is required if it is located at the
center of the site, as in this example. The costs of drilling a 30 m
well can vary widely; here, the cost is estimated at $10,000. If the
pump operates at 60% efficiency and lifts water to a total dynamic
head of 60 m, it will cost $8,000 a year to keep the pump running,
assuming that electricity costs 5C 1 kwh.
  Pumping will  have to continue for nearly 5 years to reduce con-
centrations below the  drinking water standard (Fig. 7). The total
cost, then,  would be  about $50,000, excluding the cost  of water
treatment or dilution  and labor  and analytical costs (additional
costs will depend on the specific problem). In this case, installation
of a purge well,  followed by natural restoration, may in fact be the
next best remedial action available.
Slurry Trench or Wall
  A slurry trench constructed around the site to contain the nitrate
would need to be at least 20,550 m2.  Slurry trench costs vary from
$40 to $250/m2 at medium depth." Assuming a cost  of $100/m2 of
trench area, the cost would be at least $2,000,000. If only a slurry
wall downgradient  is  used to control the  nitrate escape from the
site, the cost would be about $750,000. In addition, monitor wells
will be required  to detect any potential leaks. These costs are only
60-80% of the total costs to be incurred to complete  the slurry wall
or trench.
  Other pumping schemes are available. For example, if water is
pumped upgradient of the site so that the hydraulic gradient is
reversed near the site, the nitrate would not move out of the site, or
would do so very slowly.  This course of action may conflict with
the rights of other groundwater users and, of course, it would not
restore the site. Other options may  include artificial  recharge  or
some combination of these choices.
Ammonia

  The authors have assumed that ammonia will leach to ground-
water with  a source half-life similar  to that of nitrate. Ammonia
leach rates  for half-lives of 2 and 10 years are shown in Figure 8.
Ammonia leach  rates are twice as large as those for nitrate since the
initial amount in the ground is twice as much (200,000 kg of am-
              1600
                                                                       80,000
                       1234
                         Pumping Time (years)

                           Figure 7
  Distance from Site Boundary Beyond Which Nitrate Will Not Exceed
            10 ppm at N vs. Purge Well Pumping Time
                                                             60,000
                                                          I  40.000 -
                                                             20.000 -
                                                                                 Total Ammonia - 200.000 kg
Source Hall Life
— 2 years
   10 years
                                                                                        12      16
                                                                                       Time (years)

                                                                                    Figure 8
                                                           Amount of Ammonia Leached per Year into Groundwater System


                                                         monia compared to 100,000 kg of nitrate). The maximum concen-
                                                         trations is plotted as a function of distance in Figure 9; also shown
                                                         is the time at which maximum concentrations will occur. Note that
                                                         ammonia (Fig. 9) travels much more slowly  than nitrate (Fig. 3).
                                                         This difference occurs because ammonia  is adsorbed and therefore
                                                         retarded, while nitrate does not adsorb.  A value of one was used
                                                         for the distribution coefficient in Figure 9. When travel times are
                                                         large compared to source half-lives,  maximum concentrations are
                                                         independent of source half-life more than 1500 m  (about a mile)
                                                         from the site. The effect of dilution due to infiltration has not been
                                                         included in this analysis.
                                                           The effect of the distribution coefficient (kd) on  ammonia con-
                                                         centrations 1000 m from the center of the site is shown in Figure 10.
                                                         As kd increases, maximum concentration decreases.  The higher the
                                                         distribution coefficient, the more ammonia is adsorbed on the soil;
                                                         thus, less ammonia will appear in the ground water. A high distribu-
                                                         tion coefficient also results in high  retardation; travel time is in-
                                                         creased,  allowing for greater dispersion and, hence,  a wider plume.
                                                         Concentrations are very sensitive to the value of kj used (Fig.  10).
                                                         The  distribution coefficient  is, therefore,  a very  important
                                                         parameter in studying the spread of a contaminant.
                                                                        500
                                                                                  WOO        15OO
                                                                                Distant* Along Ctnterhne (m)
                                                                                                                 »00
                                                                                    Figure 9
                                                           Maximum Ammonia Concentrations as a Function of Distance and
                                                                         Ammonia Leach Rated (half-life)
166
CONTAMINATED GROUNDWATER CONTROL

-------
       0   20   40   60  80  100   120  140  160  180  200  220  240
                             Time (years)

                           Figure 10
 Effect of Distribution Coefficient (kj) on Ammonia Concentrations at
             1000 m from Center of Site (t,A = 2 years)
     -25O    0    250   500   750  1OOO   1250   1500  1750  2000  2250
                        Distance Along Center-line (m)

                           Figure 11
Downstream Concentrations for Uniform and Nonuniform Distribution of
 200,000 kg of Ammonia Leaching into Groundwater with t1/4 = 2 years
                          and kd =  1
            1800
            1600
                      Source Half Life
                      	2 years
                         10 years
                      4     8      12     16     20
                        Pumping Time (years)
                                                                     The effect of nonuniform distribution of ammonia over the site
                                                                   is shown in Figure 11. Again, concentrations very near the site are
                                                                   dependent on the ammonia distribution within the site, while con-
                                                                   centrations away from the  site are not sensitive to  the relative
                                                                   distribution  of ammonia. The amount  of ammonia from  each
                                                                   region in Figure 1 and the  relative concentrations  used for
                                                                   nonuniform distribution in Figure 11 are given in Table 2. (Figures
                                                                   9 and 10 are for uniform distribution of ammonia over the site.)

                                                                                               Table 2
                                                                         Amount of Ammonia in Various Regions for Nonuniform
                                                                                       Distribution in Figure 11
Region in
Figure 1
1
2
3

Amount of Ammonia
(kg)
81,325
63,253
55,422
Total: 200,000
Concentration
relative to region 1
1
3
0.57

                           Figure 12
    Distance from Site Boundary Beyond which Ammonia Will Not
         Exceed 10 ppm as N vs. Purge Well Pumping Time
Remedial Actions

  No drinking water standards have been established for ammonia.
However, a standard  of 0.02  mg/1 as un-ionized ammonia for
freshwater aquatic life has been established.13 For a temperature of
15 °C and a pH of 6.5,  the concentration can be as high as 19 mg/1
as N. For a pH of 7.5, this value is reduced to 2 mg/1. Hence, if the
groundwater flows to  a nearby stream or  river,  these pH- and
temperature-dependent ammonia standards must be  considered
together with the natural dilution provided by the stream.
  In this example, using the maximum allowable concentration of
10 mg/1 as N (based on the aquatic life standard), Figure 9 indicates
that a problem may exist for any down-gradient users within  1600
m of the site. Again, three options can be considered:

Natural Restoration
  Take no remedial action, and relocate  down-gradient users or
face possible liability.  The financial consequences of this option
will vary greatly depending on the situation. As with nitrate, a com-
bination of purge wells followed by natural restoration may be the
most viable option.

Purge Well
  Install a purge well to contain all the ammonia within the  site,
product ammonia with water and dilute or treat the water before
discharging it. Depending on the source half-life (2-10  years), the
purge well will need to be in operation for 5-20 years (Fig. 12),  pro-
ducing about 1.1 mVmin at a cost of $8,000 a year. Thus, the total
cost could be $50,000-$ 170,000 in  addition to dilution or water
treatment costs. This option still looks favorable compared to the
slurry trench.

Slurry Trench
  As discussed for nitrate, constructing a slurry trench to contain
ammonia within the sited will cost millions of dollars.
  The model can also be used to evaluate other remedial actions.
 CONCLUSIONS

   A site owner confronted with a contamination problem has a
 number of alternative remedial measures of varying cost and effec-
 tiveness to mitigate his problem. These include, in the most prob-
 able order of increasing expense:
 •Natural restoration
 •Pumping up-gradient to stabilize the contaminated water
 •Purge well  systems to contain and remove contaminated water;
 the water is either diluted or treated and discharged on the sur-
 face
 •Complete and partial slurry trenches with clay capping
 •Active restoration including chemical flushing
                                                                      CONTAMINATED GROUNDWATER CONTROL
                                                                                                                              167

-------
  The cost of these  options will depend on the site. If the con-
tamination is very localized, slurry trenches can be the most cost-
effective method. However, if the material at the base of the slurry
trench is not continuous and of sufficiently low permeability, then
contamination will still leak to the groundwater system,  and the
first three options must be given serious consideration. The authors
primarily address the natural restoration option with the purge well
method. This purging  might reduce contamination  under the site
sufficiently to let natural restoration take over.
  If only natural restoration is employed, the drinking water stan-
dard for  nitrate will  be  exceeded downstream from the site.
However,  for both assumed leaching rates this can be avoided by
using a  purge-well system for  4-5 years. As shown in  Figure 7,
pumping can be stopped after 5 years, and natural restoration  will
be adequate to  prevent serious contamination in the future and to
effect site  restoration.
  Ammonia  requires more careful consideration:  first,  because
there is generally more of the species in a fertilizer  plant, and  se-
cond, because it is a retarded species which allows concentrations
to build up under the site. Nitrate takes 4-7 years for the peak con-
centration to move from  under the site.  Ammonia, on the other
hand, takes 15-25 years, owing to retardation on clays. Since it re
mains under the site so long, leachate arriving in the aquifer con-
tinues to increase in  concentration  until the plume moves  off-site.
  For a 2-year source half-life, approximately 4 years of purge well
pumping are sufficient to make the transition to natural  restora-
tion. The  10-year half-life requires careful consideration of the im-
pact on nearby streams and water  supplies. If 10 mg/1 (as  N) is an
acceptable concentration,  then the point of impact must be slightly
greater than  1 km from the site to avoid contamination and  still
allow shutting down the purge well system.
  Other scenarios can be considered,  especially if the data show
considerable uncertainty. The computer modeling approach is vir-
tually the  only method available  to estimate the impact  of con-
taminants  and  assess the potential liability  and effectiveness of
remedial measures.

ACKNOWLEDGEMENTS
  The program described in this paper is available for PCs, as well
as for mainframe computers, under the proprietary name PLUME,
from In-Situ, Inc. This paper was presented,  under a different title
and  with   a somewhat  different emphasis,  at  The  Fertilizer
Institute's Environmental  Symposium  "84 in Orlando, Florida, in
October.

REFERENCES
  1.  Bumb,  A.C.,  Drever, J.I. and McKee, C.R.,  "In Situ Determination
    of Dispersion Coefficients and Adsorption Parameters for  Contam-
    inants Using a Push-Pull Test," presented at the Second Int. Conf. on
    Groundwater Quality Research, Tulsa, OK, Mar., 1984.
 2.  Drever, J.I.  and McKee, C.R.,  The Push-Pull  Test: A Method of
    Evaluating Formation Adsorption  Parameters for Predicting the En-
    vironmental  Effects  on  In Situ  Coal  Gasification  and  Uranium
   Recovery, U.S. Dept.  of Energy contract report DOE/TIC-11383; also
   In Situ, 4, 1980, 181-206.
 3.  Freeze, R.A. and Cherry, J.A., Groundwater, Prentice-Hall, [ingle-
    wood Cliffs,  NJ, 1979.
 4.  Way, S.C., el al., Hydrologic Characterization of Coal Seam:, for
    Methane  Recovery—Activity 2  Topical  Report: Hydrologic (. ~on-
    straints and Test Procedures; In-Situ Inc., Laramic, WY; final report
    for Gas Research  Inst.,  Chicago, IL; Contract 5082-214-0729. in
    press.
 5. Way, S.C. and McKee,  C.R., "In  Situ Determination  of Three-
    Dimensional  Aquifer Permeabilities."  Groundwater,  20, 1982, 594-
   603.
 6.  Anderson, M.P., "Using Models to Simulate the Movement of Con-
    taminants through Groundwater Flow Systems," CRC Critical Re-
    views in Environmental Control, CRC Press,  1979, 97-156.
 7.  Way, S.C. and McKee, C.R., "Restoration of In Situ Coal Gasifica-
    tion Sites  from  Naturally Occurring  Groundwater  Flow  and  Dis-
    persion," In Situ, 5,  1981, 77-101.
 8.  Fried, J. J., Groundwater Pollution (Developments in Water Science v.
    4), Elscvier, Amsterdam, 1975.
 9.  Gelhar,  L.W. and  Axness,  C.J.,  Stochastic Analysis  of Macro-
    dispersion  in  Three-Dimensionally  Heterogeneous Aquifers,  Rept.
    H-8, Hydrologic Research Program, New Mexico  Inst. of Mining &
    Technology, Socorro, 1981.
10.  Lallemand-Barres, A.  and Peaudecerf,  P., Recherche des relations
    entre la valeur de la dispersivite macroscopique d'un milieu aquifere,
    ses auires caracteristiques el les  conditions de mesure,  Bulletin du
    BRGM, 2e Serie. Section III, no.  4, 1978.
11.  Princeton University, Water Resources Program, Groundwaler Con-
    tamination from Hazardous Wastes, Prentice-Hall, Englewood Cliffs,
    NJ, 1984.
12.  Cupla, S.K., Tanji, K.K. and Luihin,  J.N., A  Three-Dimensional
    l-inite  Element Groundwater Model,  California  Water  Research
    Center, Univ. of Calif., Davis, Contrib.  152, Nov.  1975.
13.  USEPA,  Quality Criteria for  Water, Office of Water and Hazardous
    Materials, Washington, DC, July  1976.
14.  USEPA,  Slurry Trench Construction for Pollution Migration Con-
    trol, Office of Emergency  and  Remedial  Response, Washington,
    DC. EPA-540/2/84-OOI. 1984.
15.  de Josselin de Jong, G., Dispersion of a  Point Injection in an Aniso-
    Iropic Porous Medium, Geoscience Dept., New Mexico Inst. of Min-
    ing & Technology, Socorro, NM.  1972
16.  Bear, J., Dynamics  of Fluids in  Porous Media, American Elsevier,
    New York, NY. 1972.
17.  Drever, J.I., The Geochemistry  of \atural Waters,  Prentice-Hall,
    Englewood Cliffs, NJ, 1982.
                         APPENDIX A:
                         Source Functions
Instantaneous Source
  cs(t)
  When M0 mass of the contaminant reaches groundwater instantaneously
over an area A, the source function, C,, is given by
           MO
          ——     »(t)                                     (A-l)
          fAbR
where  is porosity, b is the saturated thickness, R is the retardation factor
defined by eq. 1, and 6(0 is the delta function, which is zero except at t = 0,
when ii is equal to  unity.
Constant Rale of Contamination

  The source function for a constant rate of contamination between time tt
and time 12 is given by
   C  (t) •
   5
            »AbR(t2-t,)
            0              otherwise
                                                            (A-2)
where Cs(t) is the rate of contamination given on concentration and MO is
the total amount of contaminant to appear in the groundwater at constant
rate between time t, and time U.
Exponentially Declining Contamination Rate

  It is assumed here that the rate of leaching of contaminant (M) is propor-
tional to the total amount (M) present at the time, i.e.,
   M •  kM.

When eq. A-3 is integrated, one obtains

   M -M  e-^^o).
                                                            (A-3)
                                                            (A-4)
where M0 is the total mass present in the ground at time IQ.
  The numerical value of k as well as M^ (total amount in the ground) can
be estimated either by fitting various sets of production data over time to
eqs. A-3 and A-4, or by comparing production in a particular year to the
measured amount in the ground at that time. It is assumed that all the con-
168       CONTAMINATED GROUNDWATER CONTROL

-------
 a™1"J?'lts released to the aquifer were produced, say, by the purge well,
and did not escape the contaminated area. The authors recommend use of
tne relative least-squares method to fit the data because it has the advantage
ot attempting to optimize the fit to maintain the same average percentage
error over the range of data. This makes the fit better at lower production
values than the standard least-squares approach, which minimizes absolute
error. Thus the predicted  values beyond the range of data should  be
somewhat better.
  The k value is related to the half-life (t,/2)  of the source by the equation
  t      0.693
  tl/z   ~^~ '                                             (A-5)

The source half-life is the time required to reduce the amount of contami-
nant in the ground by half,  or equivalently, the time required to reduce the
total amount of contaminant released to half the original. A typical pro-
duction curve from a purge well is shown in Figure 2.
  The source function in terms of concentration may be written as

  C  (t     --  e
  s       4>AbR
          -°-  e-Mt-t0)
                                                               (A-6)
                          APPENDIX B:
                        The Computer Model

  The basic equation for the transport of a solute in saturated porous
media can be written as:

               ^^irTxT""*                           (B-D
                 i
where
  ^- =  <  V.  >
3Vxj
  c      = concentration [M/l3]
  t      = time [T]
  V;     = mean value of ith component of fluid velocity [L/T]
  x;,Xj   = space coordinates [L]
  Dy     = i,j component of hydrodynamic dispersion tensor [L2/T]
  Rx     = rate of conversion or adsorption of solute

The general solution was derived by G. de Josselin de Jong'5 using pro-
bability theory. Effects of adsorption are accounted for by introducing a
retardation factor,  R,16-7 as defined by  eq. 1. The retardation factor con-
cept is valid for dilute concentrations."
  The specific  solution for instantaneous contamination  from a rec-
tangular site (XQ by y0) centered at the origin was given by Way and McKee.6
  C(x,y,t) =•
                 [erf
                        x+x0/2
                                 Vt/R
                            4DLt/R
                                     -J - erf (-
                                               x-x0/2
                                                        Vt/R
                                                   4DLt/R
               [erf  (-
                     y+y0/2
                         4DTt/R
                                       erf
                                           ,y-y0/2
                                                               (B-2)
                                                4DTt/R
                                                                        Eq. B-2 is valid for a uniform groundwater velocity field. Vx and Vy are the
                                                                        groundwater velocity components in the x and y directions, and DL and DT
                                                                        are longitudinal and transverse dispersion coefficients. The sides of the rec-
                                                                        tangle are parallel to the x-axis or y-axis, and the x-axis coincides with the
                                                                        direction of longitudinal dispersivity.
                                                                          For a constant strength contamination source, for a time period of T
                                                                        from a rectangular site (XQ by y0) centered at the origin, the concentration of
                                                                        solute at any time, t, and space is given by


                                                                          C(x,,.t)-Ł!L/T'[erf Ł*>"    ^(^]'\   .-, /-'"'^V<-<>/".
                                                                              •[erf (-
                                                                                                  / 4DL(t-T)/R
                                                                                             Vv(t-T)/R        v
                                                                                            -X-	) - erf (-
                                                                                                                -)   erf
                                                                                                                           / 40L(t-i)/R
                                                                                                                                       (B-3)
                                                                                                                 4DT(t-i)/R
                                                               40T(t-t)/R

                                               where

                                                 T1  = Min  [t, T} .

                                                 For an exponentially decaying strength of contamination, the concentra-
                                               tion of solute at any time, t, and space is given by
                                                                           C{x,y,t)  = —T  exp (-0.693 T/II/Z)
                                                                                       4  o
                                                                                     x+x0/2  -  V  (t-r)/R          x-x0/2 - Vv(t-T)/R
                                                                              •[erf (	=1=	)   erf  (-
                                                                                          4DL(t-r)/R
                                                                                                                     4DL(t-t)/R
                                                                                                                                       (B-5)
                                                                               •[erf  (-
                                                                                         4DT(t-t)/R
                                                                        where source strength decays according to
                                                                                               V  (t-xJ/R          y-yz -  V (t-t)/R
                                                                                             —i	)    erf  (	i—:	)] di,
                                                                                                                     4DT(t-i)/R
                                                                                    -0.693t/t
                                                                                             1/2
                                                                                                                                       (B-6)
                                               i.e., every t,A years source strength reduces to half of the original. Integra-
                                               tion in eqs. B-3 and B-5 is carried out numerically in the computer model.
                                               Other types of source functions can be handled similarly. If more than one
                                               rectangular site is of concern, then concentrations resulting from each site
                                               are added, to obtain the final concentration at a point. A mass balance of
                                               99.99% was observed for test problems, giving us confidence in the
                                               numerical integration, mathematical formulation, and the  computer
                                               model. First-order conversion of one species to another is easily handled by
                                               introducing a factor similar to "exp (-0.693 t/tc1/2)" in eqs. B-2, B-4, and
                                               B-5, where t
-------
           CONTROL TECHNOLOGY USED IN  AN AQUIFER
                     CONTAMINATION CRISIS  SITUATION

                                              JONAS A.  DIKINIS
                                    U.S. Environmental  Protection Agency
                                                 Chicago,  Illinois
                                             KENNETH J. QUINN
                                            Warzyn Engineering Inc.
                                               Madison, Wisconsin
                                             WILLIAM D. BYERS
                                                   CH2M HILL
                                                Corvallis,  Oregon
INTRODUCTION

  The Verona Well Field provides potable water to 35,000 of the res-
idents of the City of Battle Creek, MI, part of the water supply re-
quirements for two major food processing industries, the total
water supply requirements for another food processing industry
and a variety of other commercial and industrial establishments.
  During August, 1981, while conducting routine testing of private
water supplies, the Calhoun County Health Department discovered
that the water supply from the Verona Well Field was slightly con-
taminated with Volatile Organic Compounds (VOCs). Followup
testing  by the Calhoun County Health Department and the Mich-
igan Department  of Public Health (MDPH) revealed that ten of
the city's 30 wells contained detectable levels of VOCs. The MDPH
then began weekly sampling of the well field. During that same
period, the MDPH began sampling private residential wells in the
area to the south of the well field.
  Approximately  80 private wells were found with varying con-
centrations of contaminants. Several of the private wells had VOC
contamination levels up to l.OOOfjg/l, with one well having a dich-
loroethylene  concentration of 3,900 ug/1. A bottled water pro-
gram was implemented for the area residents during the time a
water supply system was being constructed to provide city water
to the affected area. The system is now complete and the bottled
water program has been discontinued.

Problem Analysis

  During the course of hydrogeologic investigations to identify
and characterize  the sources and  extent  of contamination, the
MDPH continued sampling wells within the Verona Well Field. In
the period between September, 1982 and January, 1984, the water
quality of many  wells deteriorated  and  additional wells were
affected.  Through January, 1984, blending  water  from  the least
affected wells with higher concentration  wells kept distribution
system water below the 10~' cancer risk level. Projections for in-
creased pumping  and potentially increased concentrations in the
summer of 1984 indicated that blending water within the well field
could not reduce the risks below the 10"' cancer risk level.
  To avoid  the potential for reaching unacceptable  levels in the
drinking water supply,  an Initial  Remedial Measure (IRM)  was
studied through a  Focused Feasibility Study' and was implemented
before the increased summer pumpage caused further deteriora-
tion of water quality.
  Seven compounds at various concentrations have been identified
repeatedly in the well field. These are listed in Table 1 with the an-
cillary detected compounds and available health criteria.

SITE DESCRIPTION

  The Verona Well Field is located in the northeast corner of the
City of Battle Creek. The well field consists of three wells west of
the river in Bailey Park and 27 wells and  a major pumping  and
                                                   water treatment station east of the Battle Creek River. The three
                                                   wells developed west of the river are connected to the Verona Well
                                                   Field by a pipeline under the river. Land use in the vicinity of the
                                                   Verona Well Field is light to heavy industrial with a residential area
                                                   to the south and the Grand Trunk Western Railroad marshalling
                                                   yard directly east of the field.
                                                     The monthly Average Daily Demand (ADD) for water produced
                                                   at the Verona Well Field was between 9.7 mgd and  1; mgd for the
                                                   last 2 years. The Maximum Daily Demand (MDD) during the prev-


                                                                             Tible 1
                                                                     Contaminant Characteristics
                                                   Contaminant
Higest Observed   Long-Term   Cancer RUk
 Concentration Health Advisor)   Level
   (uq/V)	(uq/lj     1.0-6 (ug/l)
                                                    Frequently Detected

                                                    1,1 Olchloroethane (1,1  OCA)    34

                                                    1,2, nichloroethane {1,2 OCA)     8

                                                    1,1,1 TMchlorethane           150
                                                    (1,1.1 TCA)

                                                    c1s-l,2 Dlchloroetnylene       229
                                                    (ds 1,2, DCE)
                 1.000
                            0.6

                            22
1,1 Dlchloroethylene (1,1 IKE)
Trtchloroethylene (TCE)
Perchloroethylene (PCE)
Sporadically Detected
Metnylene Chloride*
1 ,2 Ofbromoethane"
Chloroform*
Benzene*
Ethyl benzene*
Tol uene*
Xylene*
1,2 ntchlorobutane*
Vinyl Chloride*
trans-1 ,2-d1chloroethylene*
11
62
94

5.5
3
15
9
5
57
26
detected
1
detected
0.034
75 2.8
20 0.9

150 0.19
0.055"
0.19
70 0.67
-
100"
-
-
1.0**
-
                                                    ' Sporadic, generally nonrecurring contaminants observed in some Verona wells.
                                                    •• These levels were established by the National Academy of Sciences. All others were taken from
                                                      the USEPA and ihc Cancer Assessment Group.
170
CONTAMINATED GROUNDWATER CONTROL

-------
lous two years, 1982 and 1983, was at 18.8 mgd. As expected, the
historical demand increased through the months of May, June and
July with the MOD of 18.8 mgd  occurring in July of 1983. The
ADD in 1983 was similar to that in 1982; however, a comparison of
MDD for both years indicates much higher maximum daily usage
in 1983 than in 1982.
Hydrogeology

  The Verona Well Field is located adjacent to the Battle Creek
River in  an area mapped  as glacial alluvial deposits.2 These de-
posits vary in thickness from 8 to 77  ft in the vicinity of the well
field. The alluvial depolsits are in direct connection with the Battle
Creek River.  The well field causes a substantial amount of induced
infiltration from the river. The area outside the alluvial valley is
underlain by glacial outwash deposits.2 Underlying the glacial de-
posits is the Marshall sandstone formation. The sandstone is a fine
to medium well cemented sandstone and varies from approximate-
ly 70 to 130  ft in thickness in the vicinity of the Well Field. The
base of the sandstone  grades into the Cold Water  Shale which is
the lower boundary of the aquifer tapped by the Verona  Well
Field. Each well of the well field is cased through the glacial de-
posits (18 to 77 ft) into the Marshall Sandstone, with an open hole
in the rock.  The depth of casing into the rock varies but is gen-
erally about 10 ft. Water transmission through the Marshall forma-
tion appears to occur primarily through fractures in the sandstone.
The overall hydraulic  conductivity of the aquifer is estimated to
be approximately 1x10"' cm/sec.

PUMPING AND CONTAMINANT HISTORY

  The contaminants at the Verona Well Field were first identified
in the southern and eastern wells in September, 1981. Pumping at
this  time was centered in  the southern wells. Figure  1 is a  water
table map based on wells installed by the USEPA Technical Assis-
tance Team (TAT) and United States Geological Survey (USGS)
the year after the contamination was identified. This map indicates
that the cone of depression was centered in the southern and cen-
tral portion of the well  field.
  The extent of contaminant spread shortly after discovery of the
contamination (January through April, 1982) is shown in Figure 2.
The total VOCs shown on this drawing represent the sums of the
volatile organic compounds identified in each well. Three zones of
contamination are shown:  a zone with  1 to 50 ug/1 total VOCs,
a zone with total VOCs in  the range from 50-100 ug/1 and a zone
showing concentrations greater than 100 ug/1. Inspection of Figure
2 shows the contaminants were confined  to the southern and east-
ern portions of the well field in early 1982.
  As the southern and eastern wells became more  contaminated,
pumping was shifted to clean wells further north in the field. The
cone of depression from pumping measured in all available mon-
itoring wells on Feb. 1, 1984 is illustrated in Figure 3. Comparison
of Figure 1 with Figure 3 illustrates the shift in pumping from the
south to the north between August, 1982 and February, 1984.
  Concurrent with the shift in pumping to the north, the contam-
inants also migrated to the north. The  extent of contamination
in January, 1984 is shown in Figure 4. Comparison of this map with
Figure 2  illustrates  the contaminant migration  further north into
the well field. Closer inspection of the VOC levels in each well on
Figures 2 and 4  indicates that contamination at several wells had
increased  two- to  three-fold during  the 2-year period.  For ex-
ample, concentration of VOCs in well V35 had increased from 173
ug/1 to 343 >ig/l total VOC. Wells in the interior and northern
portion of the well field that previously had levels below  detec-
tion limits now had levels ranging from 2 ug/1 (well V37) to 45 jig/1
total VOC (well V38).

ENDANGERMENT ASSESSMENT

  As mentioned previously, the City has been able to provide water
of acceptable quality by blending contaminated well water with pri-
marily uncontaminated water. However, as the contaminant plume
 KITES:

 i. WATER LEVELS OBTAINED FROM USGS.
 1  HATCH LCVCLS AVAILABLE ARE WTEO AWACEHT TO WELL.
• II
                                4 WATER TABU CONTOURS ARE DASHED WHERE INFERRED.

                                3. WtlL LOCATIONS ARC APPROXIMATE .AND HATER TABLE
                                  HAP IS FOB PRELIMINARY USE OHLT.
                                6. BASE HAP ADOPTED FROM USGS PLAHIMtTRIC HAP.
             LOCAIIM AM KUMH OF tKU 1UIALUB IT U.I. UOLKlCAt lUtrtl
             LOCATIOt MO NIMH OF Mil IHItULIO II tCOlKT AW [MIROMIUI. IMC. HCMICAL
             ASIISTMCt KAN (IAT)
           s sras/B ysffsssraSt- ""
          • N  COUIVATIDH Kill W STATE V HlUlbM nOKITT
          A   STUM CAGIM STATION. CCMUMKM
          A   1TUM HEAJIMIHC IIU. HllaiUUICOUS
          (^ LOCATIM V TKMU ULVtHIS rACIUUES
                 IC HUE LOWIU, A[ AUK, KITH DEW HIE CWO
                            Figure 1
                        Water Table Map
advanced into the well field, blending alone did not appear to be
sufficient to maintain adequate water quality.
  As  part  of the Focused Feasibility Study, an endangerment
assessment was conducted to determine the risk that water from the
well field would present. This assessment considered the human
health risks presented by supplying the peak demand using the least
contaminated wells in the field. The human health risks were con-
sidered to be the sum of the cancer risk levels posed by individual
carcinogenic or suspected carcinogenic compounds. The cancer risk
level under these conditions in January, 1984 was estimated to be
5 x 10~'. The endangerment assessment  went on to consider the
potential human  health risks under conditions expected in the
summer of 1984 with no remedial action.  The cancer risk level was
conservatively estimated to become about 4 x 10~5.
  This endangerment assessment showed that continued deteriora-
tion of the well field would pose an unacceptable threat to human
health. The trend of the plume, as described above, indicates that
contaminant  levels would most likely increase. As a result of the
above risk calculations and the historical  trend, the authors found
the "no action" alternative unacceptable.


   CONTAMINATED GROUND WATER CONTROL       171

-------
                 : '    ' •'  ' '  •?ŁŁ&•&» •
                 >    ,--   r/Ł8sff/~
                     ' ,  v^_     c>
                    ..',  :     ""' •-;' ".-:  I-   :  •-
             \r    \  i:  L'-fc-A
                               IV
                                  •-•.:    ^?.

                                   ./

 north
   COCC'IMIIOI KMCt !• *•* It «HO «U«l«T
   10 T< tCll  «0 VM.IC 1*1 CATC i Hit TMM
   ocrccnw o mi
                          Figure 2
         Contaminated Well Field Area Total VOC 1 /82-J/82
                       Verona Well Field
                                                                   I  ttltl llrli\ »t«l«(D F«X V14S
                                                                                                           l l*»IOIt*lt( M« Wll IMif
                                                                                   Figure 3
                                                                        Water Table Map February 1, 1984
                                                                               Verona Well Field
IDENTIFICATION OF ALTERNATIVES

  The need for an IRM was established to protect human health.
The following performance criteria was established to evaluate the
remedial action alternatives:
•Capable of providing safe potable water considering that all con-
 taminant source locations had not been identified
•Capable of being implemented in a short time period (less than 4
 to 6 months)
•Capable of meeting state environmental policies
•Consistent with potential long term remedies

  The Focused Feasibility Study identified and evaluated many al-
ternatives for providing a potable water supply for the city. How-
ever, to meet both the time schedule and consistency with potential
final remedies, it was decided all or a portion of the supply would
be needed from  the existing well field. Wells VI3-VI6, V37 and
V39-V43 had been  relatively free of contamination through the
date of the study, and if concentrations did not increase they could
supply water with concentrations well below  the 10~" cancer risk
levels. This capacity would meet the average daily demand for 1982
and  1983 but would fall nearly 6 mgd short  of meeting the 1983
maximum daily demand (18.5  mgd). From this information, the
following 3 alternatives were developed and considered for detailed
analysis:
(1) Treatment of contaminated water at the well field to meet the
    maximum daily demand (up to 18.5 mgd)
                                                       (2)  Development  of 6 mgd capacity in new wells  north  of the
                                                           Verona Well Field and use of uncontaminated and/or slightly
                                                           contaminated Verona Field wells
                                                       (3)  Development of 12 mgd capacity in new wells greater than tt
                                                           mile north of the Verona Well Field
                                                       Need for a Purge System
                                                         Each alternative relied on a portion of the supply from existing
                                                       uncontaminated or  slightly  contaminated wells. Therefore,  pump-
                                                       ing contaminated water from a line of purge wells near the northern
                                                       edge of the more contaminated zone was needed  for Alternatives
                                                       2 and 3 to arrest or restrict the spread of contamination. A purge
                                                       system incorporated as part  of Alternative 1 would also be desirable
                                                       to limit the volume of water needing treatment. The selection of the
                                                       municipal wells for the purge system had to consider the location of
                                                       the contaminant sources as interpreted primarily from water qual-
                                                       ity  within the well field. The progress of the source investigations
                                                       was sufficient to provide indications on  where some sources were
                                                       located but was not sufficient to identify each source or the total
                                                       distribution of contaminants throughout the aquifer.
                                                         Therefore, the wells selected for the purge system had to be lo-
                                                       cated near the northern edge of the contaminant plume in order to
                                                       intercept all likely sources. The wells selected for the purge system
                                                       were wells V20, V22, V25, V27 and V28. These wells total 1950 gal/
                                                       min in capacity and form  a line along the northern edge of the
                                                       more contaminated portion of the plume.
172
CONTAMINATED GROUNDWATER CONTROL

-------
Disposal of Purge System Water

  The 1950 gal/min of water to be pumped from the purge system
could be disposed of in several ways. Each would require different
levels of treatment ranging from no treatment to very high removal
efficiency. The alternatives considered were:
•City of Battle Creek wastewater treatment plant
•Battle Creek River
•City's domestic water supply
  Disposal to the  Battle Creek wastewater  treatment  plant was
eliminated because of limited sewer capacity in the area and time
required to assess  the potential impact of contaminated water on
the treatment plant performance.
  Direct discharge of VOC-contaminated water to the Battle Creek
River would  not meet discharge  requirements established by the
MDNR. The state requires that discharges to surface water must
meet   Best  Available   Treatment  Economically  Achievable
(BATEA).  For this situation, BATEA was defined by MDNR as
95% removal of VOCs using air stripping. Costs estimated for the
air  stripper (Table 2) show the capital costs in year 0 and opera-
tion and maintenance (O & M)  costs over a 5-year period. The air
stripper exhaust air is considered a source of VOCs by MDNR and
was required  to have 90% of the VOCs removed from the air emis-
sions. The design influent, effluent, degree of removal and uncon-
trolled air emissions are shown in Table 3. Because of the MDNR
requirements, costs for the air  stripper included vapor phase car-
bon adsorption for the air stripper emissions.
                         (          n ^	-w    .
                 \\
                 \\  •"
^>       1>    )j
 north       /; .*-?••
                                S.V
                                                   Landfill LUUi
                                    •"   h
                                    . .        LEGEND:
                                              ran
 I.  WTCR OUHIITT IASED OH **«-«ES ".^a^lSl"1
   WtLLS KTWCE" SCPTEMCR IW) *«0 JMUttT 1M4.

 ?.  COnatltBATIW MHGE IN "'[.".rj'ErsfJJS''1
   TO IHC «IL. NO VALUt IMDICATCS LESS IHAM
   oticcnw (i ppb)
                               3.  WATER OUAIITV MW.VSES WERE OBTAINED FftOH THE CITY
                                 Of MTTIE CREEK.

                               4.  WtLL LOCATIONS ARC APPROIIHATE.

                               C.  BASE HAP ADAPTED FftON USCS PLANIMCTRIC HAP.
                         Figure 4
                Contaminated Well Field Area
                    Total VOC 9/83-1/84
                     Verona Well Field
                                                                    The present worth assumes a 5-year life, salvage value equal to
                                                                  demolition and future costs discounted at 10%. Since Michigan
                                                                  law requires use of air emission control equipment, $2 million will
                                                                  be used in the rest of this study for the 5-year present worth cost of
                                                                  air stripping as a treatment for river discharge. The alternative for
                                                                  disposal/use in the City's domestic water supply is discussed in the
                                                                  following section, Comparison of Alternatives.

                                                                  COMPARISON OF ALTERNATIVES
                                                                    Three water supply alternatives were identified above and are
                                                                  shown in Table 4. These alternatives were evaluated on the basis of
                                                                  six criteria. Each  of these criteria and results of the comparisons
                                                                  are shown in Table 4 and are discussed in the section following this
                                                                  table.

                                                                  Cost

                                                                    The 5-year present worth cost of each alternative is presented in
                                                                  Table 4. Computation of present worth was based on a 10% dis-
                                                                  count rate and neglected any salvage value. Costs of each alterna-
                                                                  tive include those for pumping and treating of the  purge system
                                                                  water ($2,000,000).
                                                                  Alternative 1
                                                                    The cost for alternative 1 assumes:
                                                                  •13 mgd is available from slightly or uncontaminated wells which
                                                                   would not require treatment, the remaining 6 mgd (4200 gal/min)
                                                                   would come from treated sources
                                                                  •25% of the time, city demands would be  supplemented by water
                                                                   produced by the purge system (1950 gal/min). The water treated

                                                                                             Table 2
                                                                                   Air Stripper Cash Flow Table

Year
0
1
2
3
4
5
Total Cost
Present Worth
Air Stripper w/o
Emission Control
$464,000
70,000
70,000
70,000
70,000
70,000
$814,000
$729,000

Air Stripper with Emission Control
$ 675,000
442,000
442.000
442,000
442,000
442,000
$2,885,000
$2,350,000
                                                                                              Table 3
                                                                         Predicted Operating Performance of Air Stripping Column
Design
Influent*
Concentration
Compound ug/1
1,1 OCA
1,2 OCA
1,1,1 TCA
cis 1,2 DCE
1,1 DCE
TCE
PCE
Total
38
8
150
229
11
62
94
592
Predicted
Effluent
Qual 1 ty
ug/1
2
6
14
2
—
2
_3
29
Expected
X Removal
(94%)
(30%)
(91%)
(99%)
(99%)
(97%)
(97%)
(95%)
Uncontrolled
Air Emission
itig/m3
1.8
0.1
6.8
11,3
0.6
3.0
4.6
28.2
Ib/day
0.8
0.1
3.2
5.2
0.3
1.4
2.1
12.9
                                                                   •Highest Concentration of each contaminant observed to date in any municipal well.


                                                                     CONTAMINATED GROUNDWATER CONTROL       173

-------
 by air stripping would have to be polished to 99% removal using
 carbon adsorption
•5% of the time, 2500 gal/min of more contaminated water would
 have to be treated using activated carbon adsorption
  A carbon usage rate of about 2 lb/1000 gal would achieve an
additional 95% removal on the treated purge system water at the
effluent concentrations shown in  Table 5. Treatment of the un-
treated higher concentration water (5% of the time at 2500 gal/
min) is estimated to consume about 3 Ib of activated  carbon per
1000 gal treated. At a replacement carbon  cost of $1.50/lb, an
annual operation and maintenance cost of $700,000/year is esti-
mated. The estimated capital cost for the system including a build-
ing for year round operation is $2,900,000.
Alternative 2
  In this alternative,  new wells of 6  mgd capacity would be in-
stalled approximately one-half mile north of the Verona well field.
These new wells would be used preferentially to meet water demand
with existing municipal wells used as needed.
  Estimated capital costs for this alternative are $1,100,000. O&M
costs associated with this alternative would be negligible.

Alternatives
  In this alternative, new wells totalling 12 mgd capacity would be
installed and pumped in preference to existing wells. As a basis for
estimating, 6 mgd of capacity is located one-half mile north  as in
Alternative 2, and 6 mgd of additional capacity is  located approx-
imately one mile north of the well field.
  Estimated capital costs for this alternative are $2,800,000. O&M
costs would be negligible.

 Likelihood of Preventing Contaminant Spread
  Alternative 1 is the least likely to prevent spread of contaminant
into slightly contaminated and  uncontaminated wells  because the
alternative would require continuous  pumping of those wells  to
minimize costs of treatment. Alternative 3 would provide the great-
est chance for success since it relies the least on pumping of exist-
ing wells close to the blocking wells and thereby  gives the block-
ing wells a better chance to halt the spreading plume.

Time Required for Implementation
  The air stripping system with air emission control would require
the longest time to implement—4 to 6 months from the start of the
design. Since this is common to all alternatives, the time factor is
equal for all.
Complexity to Implement
  Alternative 1 would be the most complex to implement due to
the number  of treatment processes involved and measures re-
quired to assure suitable quality of the treated water.
Environmental Impact
  All three alternatives were judged to be equally  low  in potential
for environmental impact.
Community Impact

  No significant difference in community impact  was recognized
between the alternatives.
  Based on the information summarized in Table 4, it is apparent
that alternative 2 is  equal to or  superior in all aspects  to Alterna-
tives 1 and 3, with  one exception: the expected reliability of the
purge system  to  completely block flow of  contaminants to the
north is somewhat lower under Alternative 2 than Alternative 3.
However,  based on preliminary flow  modeling, it was estimated
that the purge wells may form a depression, with a low divide north
of the purge wells, prohibiting flow to the north.
System Implementation

  The Focused Feasibility Study (FFS) described in the first por-
tion of this report and the REcord  of Decision (ROD) identified
                            Table 4
              Comparison of Water Supply Alternatives
Description
                      Treat Water for   6 mgd new    12 mgd new
                        Potable Use    Capacity     Capacity
Cost1 (5-year present    $7,600,000    $3,100,000    $4,800,000
  worth (>  lot discount
  rate)
Likelihood of preventing  Low
  spread of contaminants
Time required for
  Implementation
             Moderate      High


4-6 Months    4-6  Months    4-6 Months
Complexity to Implement   High

Environment a' Impact      Low

Community Impact          Low
             Moderate      Moderate

             Low           Lo«

             Low           Low
'Coil at uch •llcrnilivc include* 12,000,000 coil of purge wiict ircalmcm lyiicm

Alternative 2 as the preferred alternative. During development of
the FFS, the contaminant concentrations continued to increase in
municipal wells north of the proposed purge line, jeopardizing the
potential success  of  the  proposed IRM. Therefore, the  USEPA
implemented  an Immediate Removal Action to initiate the purge
system using an aqueous phase carbon adsorption system. This
rapid installation system would enable the purge system to operate
before summer peak pumping demands. A  cone of depression
could be developed  around the purge well system during lower
pumping stress to the north; reversal of  flow of contaminants
which had already migrated past the purge well line could begin.
  The emergency  purge well treatment  system funcing and design
was initiated on Apr. 20, 1984. Operation of the system began on
May 25, 1984. The Initial Remedial Measure funding and  Focused
Feasibility Study began on Feb. 29, 1984. The Record of  Decision
(ROD) was completed by the USEPA, and the IRM was approved
for funding by the Region V, USEPA Office on May 1, 1984. Three
additional wells designed to provide 6 mgd were installed and be-
gan pumping to the  distribution system on July 9, 1984. The air
stripper with the  air emission carbon adsorption system was in-
stalled and tested  on Aug. 10, 1984 with start up planned as soon
as the emergency  carbon adsorption system tanks indicate break-
through.

SYSTEM EFFECTIVENESS

   The installed and operating system consisted  of 3  wells de-
signed to provide  6 mgd of the base water demand and a purge sys-
tem using 4 or 5 municipal wells with a temporary  carbon adsorp-
tion treatment system. At the time of  this  report (August, 1984),
the total system has been operating for approximately 3 months,
and its  effectiveness is discussed below. The permanent (5-year
expected life) air  stripper with emission control treatment system
is tested and ready to  go on line. Therefore, the effectiveness of
this treatment system has not been evaluated during daily oper-
ations.
   The effectiveness  of the 3 new wells can be measured by the
fraction of the designed  flow the wells are producing. The wells
were planned to  be placed adjacent to the Battle  Creek River to
obtain the maximum induced flow  from the river. However, be-
cause of the wetland conditions and flood  plain restrictions adja-
cent to the river,  the wells were placed further from the river. Re-
gardless of this less desirable location and modifications to the ex-
pected design, the wells are producing 85 to 95% of the  designed
production (5.1 to 5.7 mgd of the 6 mgd design capacity).
   The effectiveness of the purge system can be measured  by  the
changes inwater quality in the northern production wells (V30 and
V38 through V43) and, secondarily,  by the presence of a distinct
174       CONTAMINATED GROUNDWATER CONTROL

-------
cone of depression around the purge wells. Because the purge sys-
tem has been operating since May 25 and water quality is available
through Aug. 7, only preliminary indications  using water quality
data can be made. The change with time of total VOCs at three
wells north of the purge system, Wells V30, V38 and V39 is shown
in Figure 5. Prior  to the start-up of the  purge system (May 25,
1984), concentrations were generally increasing; after start-up of
the purge system, a noticeable downward trend occurs in each well.
The trend is admittedly short-lived at the time of this report. Addi-
tional sampling is continuing. In addition to the short-term down-
ward  trend in concentrations, a preliminary water table map for
August,  1984 (Figure 6), indicates  a cone of  depression may be
centered around the purge well  line. This distinct cone  of de-
pression around the purge wells would  prevent migration of con-
taminants beyond the purge system.
CONCLUSIONS

•The hydrogeologic  information  that  supported the  IRM was
 based primarily on the water quality monitoring of the municipal
 wells,  although additional hydrogeologic data  from  numerous
 monitor wells provided important supporting data.
•Relatively rapid contaminant migration to previously unaffected
 wells initiated the need for the IRM.
•The FFS considered several options in detail to determine the cost
 effectiveness, compatibility with potential  long term remedies
 among other criteria.
•The IRM,  ROD and emergency measure were expedited so that
 the purge system for blocking further contaminant migration was
 implemented in three months.
•The entire  IRM was operational less than 5 months after worsen-
 ing conditions dictated the need for it.
•The purge system appears to be functioning better than  antic-
 ipated in blocking continued migration of contaminants.
•After less than 2 months of operation of the IRM, the city is re-
 ceiving water with no detectable VOCs.

REFERENCES
1. CH2M HILL Inc., Focused Feasibility Study, Verona Well Field, Battle
   Creek, Michigan, Public Comment Draft, Mar. 28, 1984.
2. Vanlier, K.E.,  Groundwater Resources of the Battle Creek Area, Mich-
   igan, Water Investigation 4, Michigan Geological Survey Division,
   1966.
                                              NOTES:

                                              1  WATER LEVELS OtTAIKED FKM IMS.

                                              1. RIVER STAGE AT THE DAM Is APPRO*[MATELT 823.9
                                                ABOVE DAM AND BIB BELOW DAM.

                                              J. WATCN LEVELS AVAILABLE ARE NOTED ADJACENT TO WILL .
                                                                      *  WATER TABLE CONTOURS ARE DASHED WERE INFERRED.

                                                                      5  WELL LOCATIONS ARC APPHOIIHATE.AHO WATER TABLE
                                                                        HAP IS FOR PRELIhlNART USE DNLT.
                                                                      6.  BASE HAP ADOPTED FROM USGS PLANIHfTRIC NAP.
                                                                          Figure 6
                                                        Water Table Map, August 1984, Verona Well Field
                 o
                 o
75

70


65


60

55


50


15

4O


35


30

25


20

15

10


5

NO
                                                                                                      PURGE LIVE
                                                                                                      START UP MAY 25
                                                 E*»JULY ^UG I Sfft I  OCT I NOV ' DEC  TAN
&   WELL V30
^   WELL V38
.$-   WELL V39
A   ABOVE WELLS COMBINED
HD   NOT  DETECTED
                                                             Figure 5
                                         VOCs Graph, January 1983 to Present, Verona Well Field
                                                                      CONTAMINATED GROUNDWATER CONTROL
                                                                                                         175

-------
          CONTROL OF  ORGANIC AIR  EMISSIONS FROM
           GROUNDWATER  CLEANUP  —  A CASE  STUDY

                                          ABRAHAM THOMAS
                                     FREDERICK BOPP,  III, Ph.D.
                                              JOHN NOLAND
                                         JOHN BARONE, Ph.D.
                                            Roy F.  Weston, Inc.
                                        West Chester,  Pennsylvania
                                           THOMAS PIERSON
                                   Standard Chlorine  of Delaware, Inc.
                                         Delaware City, Delaware
                                            MICHAEL APGAR
               Delaware Department of Natural Resources and Environmental Control
                                              Dover, Delaware
INTRODUCTION
  An accidental spill of reagent grade monochlorobenzene (MCB)
occurred while workers were filling a railroad tanker car at Stan-
dard Chlorine of Delaware, Inc. (SCD1), located in Delaware City,
Delaware. An inventory analysis indicated that the volume of MCB
discharged to the ground surface could have been as much as 5,000
gal. Since the specific gravity of MCB is 1.1066, it will sink through
a soil or water column as a coherent "slug".
  The spill occurred during a rainstorm, and part of the spillage
ran off in surface ditches toward a creek running adjacent to the
property. Prompt action by plant  personnel  and  the Delaware
Department of Natural Resources  and  Environmental Control
(DNREQ (installing dams and pumps), resulted in the recovery of
most of the runoff portion of the spill.
  MCB-contaminated soils from the ditches were excavated and
disposed of under the supervision of the State DNREC. A lim-
ited test boring, soil sampling and analytical  program conducted
by the plant in the immediate area of the spill indicted the presence
of MCB to depths up to 40 ft below grade. The data gathered from
this program led the plant and regulatory personnel to the con-
clusion that the potential for groundwater contamination existed.
SCDI  retained Roy F. Weston, Inc. (WESTON) to provide tech-
nical services in defining the extent and magnitude of the problem
and in developing cost effective remedial action alternatives.
  The purpose of this paper is to report  on the successful collab-
oration of private industry, State regulatory  personnel and con-
sulting scientists and engineers in evaluating this problem and arriv-
ing at cost-effective, environmentally sound and regulatorily accep-
table solutions to the problem.
HYDROGEOLOGIC INVESTIGATION
Description of Site

  The plant lies in the Atlantic Coastal Plain of Delaware. The
surficial sands of the Pleistocene age Columbia Group1 uncon-
formably overlie the Cretaceous age Potomac and Merchantville
Formations. The Potomac Formation is underlain by bedrock. The
Potomac Formation consists of clay and silt interbedded with sand
and some gravel.2 The upper zone of the Potomac is dominantly
high-plastic clays (CH) with recorded thicknesses of up to 125 ft.
The lower zone of the Potomac is dominantly fine sand to gravel
and forms the major regional aquifer. A  search of published liter-
ature of the area indicated: (1) the presence of only two actively
pumping wells constructed in the Columbia Formation within one
mile of the spill area; and (2) the Potomac Formation is a contin-
uous confining layer beneath the plant site to a distance of about
                                                   one mile in all directions. Red Lion Creek runs along the north
                                                   side of the plant property and drains into the upper Delaware Bay.
                                                   Subsurface Exploration and Monitor Well Installation

                                                     This program was conducted in two phases. During Phase I, ten
                                                   monitoring wells were installed. Based on Phase I results, a Phase
                                                   11 program was undertaken; 21 exploratory borings and ten addi-
                                                   tional monitoring wells were drilled on or near the plant property
                                                   (Fig. 1).
                                                     All exploratory  drilling was accomplished using conventional
                                                   hollow-stem  auger equipment, and all wells were installed using
                                                   mud-rotary equipment. All monitoring wells were constructed of
                                                   galvanized iron pipe and well screens due to the incompatibility of
                                                   MCB with polyvinylchloride.
                                                     The wells  were drilled to the top of the Merchantville Forma-
                                                   tion or to the top of the Potomac clay, whichever occurred first.
                                                   The  stratigraphic logs confirmed at least 4  ft of the confining
                                                   Potomac Clay in all 31 borings at an average depth of about 70 ft.
                                                   The Merchantville Formation, with an average thickness of 13 ft,
                                                   was confirmed in  17 of the 31 borings and apparently has been
                                                   eroded away locally in the vicinity  of the plant. The results from
                                                   the test boring program indicated  that the Potomac clay would
                                                   form an effective barrier to the vertical migration of any contam-
                                                   inants  present  from the Columbia  into the  underlying aquifer
                                                   sands.
                                                     A topographic map of the surface of the Potomac and Merchant-
                                                   ville Formations is shown in Figure 1. North of the plant, a steep
                                                   erosional  gully occurs in  the  pre-Columbia surface centered on
                                                   Boring T13-41. The location of the  Merchantville Formation on
                                                   the west of the plant is also shown in Figure 1.
                                                     The configuration of the surface of the Cretaceous age forma-
                                                   tions could  influence the  direction  of migration of groundwater
                                                   contaminants.  A  cross-section through  the main  plant area,
                                                   paralleling the principal  flow direction for  groundwater in the
                                                   Columbia sand and showing  the key stratigraphic relationships,
                                                   is given in Figure 2.

                                                   Water Quality Sampling and Analysis

                                                     All monitoring wells installed at the plant were sampled approx-
                                                   imately 2 weeks after completion of well construction development.
                                                   A Kemmerer sampler was used to collect a water sample from the
                                                   bottom  one  foot interval  from each well to  obtain a worst case
                                                   estimate  of  contaminant conditions. The  concentrations of the
                                                   major benzene species are shown in Table 1. These data were used
                                                   to draw  the isoconcentration  map shown in Figure 3. The map
                                                   illustrates the distribution of total benzene species detected, or the
176
CONTAMINATED GROUNDWATER CONTROL

-------
                Legend
                 •~"  Monitor Well
                 •IS"  Test Boring
                      Spill Area
                	Merchantville (Km)/
               A—A
 Potomac (Kp) Contact
, Line of Section
 For Figure 2
                                                           Figure 1
                                   Topographic Map of Cretaceous Surface (MSL Datum). Dot Pattern
                                   Indicates Area of Merchantville FM. Subcrop. Map Also Illustrates
                                           Locations of all Test Borings and Monitor Wells
                           Table 1
 Summary of Benzene Species Concentrations in the Groundwater (mg/l)
Well No.
1
2
3
4
5
6
7
8
9
10
22
24
25
26
30
31
44
48
49
50
Benzene
ND
ND
ND
11.0
66. 8
NO
ND
206.0
166.0
107.0
ND
ND
ND
129.0
123. u
48.5
60.2
194.0
86.3
136. u
Mono-
60.0
ND
3. J
3.3
300.0
27.0
28.9
134. u
134.0
117.0
ND
18. /
10.1
176.0
107. u
17.8
58.3
118. u
132.0
94.8
o-dichloro
12.4
ND
ND
ND
144.0
32. 3
17.8
40.3
35.3
86.9
ND
8.^
8.3
57.7
36. ^
10. 3
17.6
28. /
34.3
21.6
p-dichloro
7.8
7.2
b . /
b. J
51.9
9.7
28.6
18.0
12.2
49.4
3.2
9.5
14.1
76. *
35.9
10.2
28.0
32.3
43.7
20.0
1,2,4-
trichloro
10.2
ND
/ . u
ND
163.0
10.3
7.1
9.8
3. i
24. J
5.6
9.4
7.1
23.3
12.8
7.1
5.7
/. *
11.7
5.8
«ND means not detected.
                                             row-sums from Table 1. Contaminants appear to be migrating in a
                                             generally northerly direction in a fairly cohesive plume,  and the
                                             influence of pre-Columbia topography upon that migration pat-
                                             tern does not appear to be significant.
                                             Hydrogeological Analysis

                                               Water levels  have  been  measured  in the 20 monitoring wells
                                             periodically since their construction.  The data gathered on June
                                             23, 1983, were used to draw the water table contour map shown
                                             in Figure 4;  the map shows that the dominant water table  flow
                                             direction at the plant is northerly. This is illustrated by the selected
                                             flow net rays shown on that figure.
                                               Comparison of Figure 4 with the isoconcentration map shown
                                             in Figure 3 indicates the groundwater flow direction at the plant is
                                             exerting the major influence over the observed contaminant migra-
                                             tion pattern.
                                               Of critical importance are the control of further migration of
                                             contaminants and the recovery  of these contaminants from the
                                             groundwater flow regime. Physical methods for containment of the
                                             migrating contaminants,  such as bentonite slurry walls, were  eval-
                                             uated but were dismissed from further consideration due to tech-
                                             nical difficulty in construction, integrity of the technology in this
                                             application and  costs. Since any physical barrier would have to be
                                             accompanied by recovery of contaminants at pumping wells and
                                             treatment at the surface,  hydrodynamic barrier controls were con-
                                             sidered to be the most appropriate technology to apply at SCDI.
                                                                    CONTAMINATED GROUNDWATER CONTROL
                                                                                                                          177

-------
                                                             Figure 2
                                           Cross-Section Parallel with Principal Flow Direction
                                                     (See Figure I for Location)
                                                             Figure 3
                                      Isoconcentralion Map for Total Benzene Species Detected (mg/l)
178      CONTAMIANTED GROUNDWATER CONTROL

-------
                                                                 I  A  Recovery Well Location
                                                          Figure 4
                                         Water Table, 23 June 1983. Locations of Proposed
                                                Recovery Wells Are Also Shown
  During the Phase I operations, a pump test indicated that the
transmissivity of  the  Columbia  Formation ranged from  27,500
to 66,000 gal/day/ft.  Seepage velocities  computed  from these
values ranged from about 1.2 to about 10.2 ft/day and averaged
about 3.6 ft/day.  These data were used to design a barrier control
and recovery pumping well  field for  the SCDI plant site. At  a
nominal pumpage rate of 40  gal/min/well, it was determined that
four pumping wells would be needed to control further migration
of the contaminants toward Red Lion Creek.
  The recovery wells were constructed at the four locations shown
shown on Figure 4. Each well was drilled using a  cable tool drill
and was  constructed  using an 8-itf. diameter casing  and  a tele-
scoping  8-in.  diameter  stainless steel screen. Each  well was
equipped with a  vertical turbine pump of all brass,  bronze or
stainless steel construction with cast iron bowls.
  While the design basis for the recovery well field was 40 gal/
min/week, each pump has the capacity to pump up to about 100
gal/min.  It is anticipated that  the maximum combined pumpage
rate needed to intercept and recover the groundwater contaminants
would be about 250 gal/min.  This pumpage rate provided the basis
for the evaluation and design  of the treatment facilities discussed in
the following section.
CONCEPTUAL ENGINEERING INVESTIGATION

Evaluation of Options for Groundwater Treatment

  Insufficient data were available to predict performance of the
existing wastewater treatment plant at higher than current flows
and contaminant loadings,  and certain  elements of the existing
plant (such as the wastewater clarifier) would be hydraulically over-
loaded at  an excess flow of 250 gal/min. Therefore,  WESTON
investigated  several other options  for  treatment of recovered
groundwater.
  Air Stripping. Option 1 consisted of a packed tower which was
used to air strip the benzenes out of the groundwater  down to a
level of 2.5 mg/1 in the tower effluent.  The effluent would then
proceed, as shown in Figure 5, through a 100,000-gal flow equal-
ization and stripping tank and would be split between two clar-
ifiers, one existing and one new unit, both the same size. The clar-
ified effluent would then flow through a filtration unit and then to
the effluent flume. The 100,000-gal flow equalization and stripping
tank would  provide additional  "polishing", and  the clarifiers
would remove biological and flocculated solids formed as a result
of air stripping.
                                                                  CONTAMINATED GROUNDWATER CONTROL      179

-------
                a
                       * Total benzene and derivatives
                                                            Figure 5
                                               Groundwater Treatment Flow Diagram
   The estimated emissions of benzene compounds to the air as a
 result of air stripping are  shown in Table 2.  Emissions  at these
 levels would necessitate installation of air emission controls to limit
 air emissions to acceptable levels.
   Steam Distillation.  Option 2 consisted of  a scaled-up, high-
 temperature, steam distillation system similar to that used for sol-
 vent recovery in the main plant. Estimated annual operating costs
 of this system would exceed  $1,000,000,  mainly for energy.  This
 additional  system  would consume all the plant's  standby steam
 generating capacity. Since  this situation  would be untenable for
 production, additional costs would be incurred in the construction
 of a separate, packaged boiler unit to power the steam distillation
 apparatus. Because of these excessively high costs, it was recom-
 mended that steam distillation not be considered further.
   Activated Carbon Treatment. Option 3 consisted of direct treat-
 ment of pumped effluent in granular activated carbon (GAC) col-
 umns. Based on removal of total benzenes down to a level of 2.5
 mg/1, WESTON estimated that the carbon requirement would  be
 on the order of 33,000 Id/d a yr. Assuming that this carbon could
 be steam-regenerated in-place, as is currently done at SCDI, the
 average daily carbon loss during regeneration would be  approx-
 imately 10%, or about 3,300  Ib/day  At Sl.OO/lb, the annual cost
 for carbon  replacement alone would  be  in excess  of $1,200,000.
 Based on these excessively high costs, it was  recommended that
 carbon treatment not be considered further.
   Options 2 and 3, while technically viable, were eliminated  from
 further consideration due primarily to excessively high costs.
 Option 1,  Air  Stripping with associated emission controls, was
 selected as the most efficient and cost-effective method for ground-
 watekr cleanup. DNREC concurrence in-principal with this system
 resulted  in design  of the basic treatment plant and evaluation of
 options to provide the necessary air emissions control.
ASSESSMENT OF AIR EMISSION CONTROL OPTIONS

  Since  uncontrolled emissions of benzene  compounds  to  the
atmosphere at the loading shown in Table 2 would  be unaccep-
table to  SCDI and to the DNREC, two  control strategies were
selected as technically feasible options to reduce air pollution:
•Air stripping with exhaust gases vented to a fume incinerator
•Air  stripping with  exhaust gases vented to  an existing process
 boiler
                                                          Various  alternative  refrigeration and condensation processes
                                                        were considered but were not feasible due to the low concentra-
                                                        tions of benzene compounds in the air stripping tower off-gas.

                                                        Gas Venting to a Fume Incinerator (Option 1-1)

                                                          Factors influencing the efficiency of an incinerator are tempera-
                                                        ture,  degree  of mixing and residence  time in the combustion
                                                        chamber. Only  enclosed combustion was evaluated for this option
                                                        because open flares may only be 60% efficient for benzene destruc-
                                                        tion.1
                                                          Use of both  a primary and a secondary  combustion chamber
                                                        will ensure a high degree of mixing, combustion zone tempera-
                                                        tures in excess of 1,600°F and a residence time  of not less than 1
                                                        sec (Table 3). This conservative design should achieve at least 99%
                                                        control of the chlorinated benzene compounds  since these chem-
                                                        icals have autoignition temperatures of approximately 1,200 °F.
                                                        Gas Venting to an Existing Boiler (Option 1-2)

                                                          This option is similar to the first option, except that an existing
                                                        125,000 Ib/hr process boiler would serve as the  enclosed combus-
                                                                                     Table2
                                                             Mass Emissions of Benzene Compounds from Air Stripping Tower
                                                              Compound
Itoniane
MCB1
1,2-OCB*
1,4-DCB1
TCB<
TOTAL
ll.o
10. >
4.5
I.I
2.2
14.1
126
2<2
106
74
51
• 21
».6
47.7
19.7
11.6
».»
150.2
                                                              Monochlorob«ni«n« or chlorob*ni«n*  {fornu 1*t  C.H.C1).
                                                              I,4-Dichlorob*ni«n« or p-Dlchlorobonaon* (Formulai  i,4-C.H Cl-l-
                                                                        Tr ichlorob«nt«n» (FormuU i  C.H ,CI
ISO
CONTAMINATED GROUNDWATER CONTROL

-------
                            Table3
            Design Data for Air Emission Control Options
Description
                     Option 1-1
                                             Design Data

                                         Option 1-2
Combustor Type


Type of Fuel


Burner Capacity

Combustion Zone
Temperature

Residence Time

Emission Control
System

Design Control
Efficiency
                    Two Stage, Fume
                    Incinerator

                    Natural Gas
                     11 x 10'Btu/hr

                     1,600 °F


                     1 second minimum

                     Wet Scrubber


                     VOC's — 99% +
                     HCL — 90%
B&W Model FM-117-88C,
Boiler

Nat. Gas/No. 2 or No. 6
Fuel Oil

160 x 10* Btu/hr

 2,000 "F


1 second minimum

None


VOC's — 99% +
HCL — 0%
tion chamber. At the combustion temperatures and residence time
shown in Table 3 for the Standard Chlorine process boiler, this sys-
tem is expected to achieve a benzene removal efficiency as high as
99%."
  Babcock and Wilcox, the boiler manufacturer, indicated that the
HC1 emissions would not cause a problem as long as the flue gas
temperature remained above the acid dewpoint (which is expected
to be between 300 and 3SOT). This does not appear to pose a prob-
lem since the design economizer outlet temperature is 422 °F  and
typical stack temperature is in excess of 350 °F.
Air Quality

  The estimated controlled air emissions for the two options are
summarized in Table 4. The purpose of the  air impact analysis
was to ensure that the air pollution control systems under consid-
eration are sufficient to minimize off-site ambient air impact. A
dispersion modeling analysis was used to demonstrate that the am-
bient concentrations  due to each option are below relevant health
effects criteria levels.
  Several different types of health effects can result from exposure
to the substances listed in Table 4, including: long-term body bur-
den exposure to carcinogens; long-term chronic exposure to non-
carcinogens;  short-term acute effects; and odor. The substance of
greatest concern from a long-term carcinogenic perspective  is ben-
zene.  The USEPA Cancer Assessment Group (CAG) has estimated
the exposure level which relates to a 10"! cancer risk at 0.208 ug/m3
of benzene.5  When this body burden is corrected for the expected
10-yr  project duration, as opposed to the assumed 60-yr life-time
period, the acceptable exposure is six times higher than 0.208 ug/
m3, or 1.25 ug/m3.  For the long-term cumulative effects  of the
chlorinated benzene  compounds, the best available exposure cri-
teria are the Threshold Limit  Values (TLV)  corrected by some
safety factor. This approach has been used in several states that
have attempted to establish guidelines for toxic air pollutants. A
typical safety factor which has been used in several of these guide-
line documents is  the TLV divided by 420.' Weston applied this
safety factor in this evaluation.
  HC1 is a severe irritant which can cause short-term acute effects.
The Philadelphia Toxic Air Pollutant guideline has set the same
safety  factor  on HC1 exposure  of the TLV/42.7 Weston used that
                                                             Table 4
                                          Summary of Estimated Controlled Air Emission Rates
                           Pollutant
                                                          Alternative  Control Options
                                               Option 1-1 - Air
                                               Stripping/Fume
                                               Incinerator
                                               (Ib/hr)  (ppm )  (g/sec)
                                                                         Option  1-2  - Air
                                                                         Stripping/Process
                                                                         Boiler1
                                                                         (Ib/hr)  (ppm )   (g/sec)
Particulates
(TSP)
Sulfur Dioxide
(SO,)
Nitrogen
Oxides
(NO )
CarBon
Monoxide
(CO)
Hydrogen
Chloride
(HC1)
Volatile
Organic
Compounds
( VOC ' s )
- Benzene
MCB
1,2-DCB
1,4-DCB
TCB
Other Non-
Methane
VOC's
- Total
VOC's
0.
0.
1.
0.

0.



0.
0.
0.
0.
0.


0.

0.
05
006
50
37

86



14
11
05
03
02


03

38
N/A
0.09
31. 42
12.7

22.



1.
0,
0.
0,
0,


1,

5.

.7



.72
.94
.33
.19
.11

T
,813

.10

0.108



0.017
0.014
0.006
0.004
0.003


0.004

0.048
0.
0.
0.
0.

8.



0.
0.
0.
0.
0.


0.

0.
008
001
217
054

6



14
11
05
03
02


004

354
N/A
0.003
0.872
0.36

43



0.
0.
0.
0.
0.


0.

0.

.4



33
18
06
04
02

3
05J

68

1.



0.
0.
0.
0.
0.


0.

0.

084



017
014
006
004
003


001

045
                           10ption  2  emission rates are based on  incremental increases in emissions
                            above the current boiler No. 3 emission rates.

                            NOX ppmv  concentrations are based on  the molecular weight of N02
                            (46 Ib/mole).

                           30ther VOC's  ppmv concentrations are based  on  the molecular weight of
                            CH4 (16 Ib/mole).  These VOC's result from the  combustion of natural
                            gas and are  nonhazardous.
                                                                     CONTAMINATED GROUNDWATER CONTROL
                                                                                                                             181

-------
                                                 Figure 6
                        Isopleth Diagram of Ambient Concentrations Due to Air Stripping with a
                                         Fume Incinerator for Control
                                                  Table S
                         Predicted Average Annual Concentrations (/ig/np) for Two Eratalon
                          Control Options Compared with Health Guideline Exposure Levels
odor Cfction 1-1 Option 1-2
Compound Threshhold TtV/420 Peak Receptor Critical Receptor Peak Receptor
Benzene 516 71.4 0.013 0.003 0.111
MCB 972 833. 3 0.011 0.003 0.091
1,2-DCB 305,000 714.3 0.005 0.001 0.039
1,4-DCB 90,000 1,071.4 0.003 0.001 0.026
TCB — 95.2 0.002 0.001 0.020
HCL* 1,400 156 0.814 0.201 0.702
* TLV/42
Critical Receptor
0.
0.
0.
0.
0.
0.

048
039
017
Oil
008
302

182
CONTAMINATED OROUNDWATER CONTROL

-------
                                            Actual concentrations [or each pollute
                                                aliens by the actual
                                                            Figure 7
                                  Isopleth Diagram of Ambient Concentrations Due to Air Stripping with a
                                                    Process Boiler for Control
same conservative approach in the determination of health effects
from HC1 exposure.
  All substances  listed in Table 4 were included in the modeling
analysis; in certain cases,  they  are considered to be potentially
hazardous. Criteria pollutants such as TSP (particulates), CO and
SC>2 were  not  modeled  since the controlled  emission  rates for
either option are well below the PSD (prevention of significant de-
terioration) significance level and would be well below applicable
standards.
  A USEPA-approved UNAMAP screening model (PTPLU) was
used to determine the downwind distance where peak concentra-
tions would occur for a variety of meteorological conditions. Sub-
sequently, this information was used to develop the source receptor
grid network for use in a more refined UNAMAP model (ISCLT).
The ISCLT model was used in conjunction with the 1965 through
1981 meteorological STAR distribution for Philadelphia to predict
the annual ambient air quality impact due to emissions from both
control options.
  The  Philadelphia meteorological data were used since they rep-
resent the closest and most complete data base available. A dense
polar grid coordinate network with a grid spacing of 330 ft (100 m)
to a downwind distance of 3300 ft. (1,000 m) for every  10 degrees
of arc, and with an expanded grid space of 25 ft (250 m) out to
8250 ft (2,500 m), was used  in the modeling analysis.  Since  the
height  of the existing boiler stack is less than the Good Engineer-
ing Practice (GEP) stack height, the building downwash or wake
effects option was used in the modeling analysis.
  The  results of the modeling analysis are presented in Table 5 for
the two control options. The incinerator control option would re-
sult in concentration levels which are an order of magnitude below
the CAG for benzene  at the peak off-site receptor. All other pre-
dicted pollutant concentrations for benzene compounds are several
orders  of magnitude below the TLV/420 at the peak off-site recep-
tor. The predicted concentration at the critical receptor (the near-
est residence) is a factor of 26 below the CAG value for benzene.
The TLV/420 values for the other benzene compounds are about
12,000 to 100,000 times greater than the predicted levels. The TLV/
42 for  HC1 is 200 to 500 times greater than the predicted values at
the peak and critical receptors.  The odor threshold for all pollu-
tants is also significantly greater than the predicted concentrations.
  The  boiler control option results indicate that ambient concen-
tration levels would be a factor of about 100 to 180 below the CAG
for benzene for the peak and critical receptors.  The TLV/420  is
predicted to be about 40,000 to 350,000 times greater than  any pre-

  CONTAMINATED GROUND WATER CONTROL        183

-------
dieted  peak ambient  concentrations for the chlorinated-benzene
compounds. The TLV/42 for HC1 is about 200 times greater than
the predicted ambient levels at any receptor. The predicted values
are also well below the odor thresholds for all compounds.
  To provide a better understanding of the spatial distribution of
ambient  pollutant concentrations for each option, concentration
isopleth  diagrams were developed (Figs. 6 and  7). These figures
show the concentrations which would occur if 1 g/sec of a pollu-
tant were emitted (i.e., unit emissions) by the control options.
Hence, to arrive at the concentration for any particular  pollutant,
the isoline concentration should be multiplied by the appropriate
g/sec emission rates from Table 4. For example, the 0.1 ug/m1 iso-
line concentrations should be multiplied by 0.017  (from Table 4)
to obtain the predicted  ambient annual benzene concentration for
the boiler control option isoline. The important point  to note is
how concentrations change as a function of downwind distance.
Hence, the diagrams provide a better perspective on the spatial dis-
tribution of pollutant concentrations at any relevant receptor to
appropriate health benchmarks.
   The results of the modeling analysis indicated that either of the
proposed control technologies would provide a substantial margin
of safety over any health benchmark level based on an annual ex-
posure criterion. Although the modeling analysis  used  predicted
annual average ambient concentrations, a 24-hr peak concentra-
tion can be  estimated by multiplying the average annual values by
a factor  of  10. This estimate is consistent with the  USEPA guide-
lines for dispersion modeling. Since the predicted annual ambient
concentration for any receptor or pollutant  for either control  op-
tion is at least a  factor of 20 below any applicable health bench-
mark level, it is unlikely that the emissions due to the proposed con-
trol technologies would result in an exceedance of  the benchmark
for any 24-hr period.


                            Table 6
     Description
                            Option 1 - Air  Option J • Mr
                            St r Ipptng/Fuim.*  St I lp|il nn/Pf txf
     Total Capita! Costs
      - Air Stripping System
      - Control Systesi
     TOTAL
     Annual Operating Costs
      - Air Stripping Systesi
      - Control System
     TOTAL
                  i  514,000
                  ?541,000
                                           27.000

                                           66,000
     Total 10-Year Operating Costs  5,410,000
     Total 10-year Costs         6,024,000
     Incremental Cost per pound or  $
     Benzene Compounds Removal Over
     Uncontrolled Air stripping
                                               0.18/1"
   From  an environmental point  of view, the second option, air
stripping with a process boiler for control of pollutant emissions,
is more attractive. Although this option would result in air emis-
sions, the magnitude of the controlled  emissions would be small
and the ambient air quality impact would be minimal. Comparison
of the predicted ambient concentrations due to these emissions to
all appropriate  benchmarks of public  health suggest that there
would be no adverse effects due to implementation of this option.
This condition would also be true for the first option, air stripping
with  a fume  incinerator for control. Hence,  from an  environ-
mental/health effects perspective, both options would be accept-
able.  Environmental  permitting  for chemical destruction in  the
boiler would be straightforward; however, the first would require a
significantly larger permitting effort.
   The economic impacts of the  two options are quite different.
From both  a capital and an operations and maintenance (O&M)
point of view, the most attractive option is the process boiler con-
trol option. It has lower capital and O&M costs and a lower unit
treatment cost.
  Based on the above considerations, air stripping with a process
boiler for control of air  pollutant emissions was recommended.
It would be the most economical option and would be acceptable
from an environmental permitting and public health perspective.
WESTON has recommended this option to SCDI which has, in
turn, proposed this system to the Delaware DNREC. The DNREC
has approved, and all parties are  working together to obtain the
other necessary permits so that cleanup can commence.
Recommendation of an Alternative Control Option

  WESTON has prepared an estimate of the total capital costs and
annual operating costs for each alternative control option (Table
6). To compare  the two alternatives, the total capital costs were
added to the total  estimated operating costs over the life of the re-
medial action  project (assumed to be 10 years). These total 10-yr
costs (in constant 1983 dollars) were then divided by the total mass
of additional benzene compounds removed to determine the net
unit  cost of the control option. As shown  in Table 6, the second
air pollution control option is more than five times more cost-effec-
tive than the first.

CONCLUSIONS
  As a result of spills during the history of operation of the SCDI
Plant, groundwater in the water table aquifer at the plant is con-
taminated with benzene compounds. SCDI has made the financial
commitment to undertake  cleanup of  the  water table aquifer.
WESTON has provided expertise in evaluating the magnitude and
extent of groundwater contamination and in evaluating a broad
range of  remedial action technologies which would  be  used  to
affect cleanup. A technically feasible, environmentally and regula-
torily sound, cost  effective set of control options has been recom-
mended to SCDI and  to the DNREC. The State has provided in-
valuable assistance and cooperation in working with  the consul-
tant and SCDI.
  All  parties  have agreed in principal  with  the remedial actions
recommended by  the consultant  and are currently working to-
gether to  complete the permitting effort required for implementa-
tion. This project  serves as a classic example of the  benefits to be
derived from cooperation and mutual assistance between govern-
ment and private industry in achieving environmental quality goals.

REFERENCES
1. Jordan,1976.
2. Pickett, 1970, Geology of the Chesapeake and Delaware Canal Ana,
   Delaware, Delaware Geologic Survey,  Geologic Map Series, No. 1,
   Scale—1:24,000.
3. "USEPA, Proposed National Emission Standard for Hazardous Air
   Pollutanis—Benzene  Fugitive Emissions",  Environmental Reporter,
   Jan. 9, 1981.
4. "USEPA Proposed Standard for Benzene Emissions  for Maleic Anhy-
   drite Plants", Environmental Reporter, Apr. 18,1980.
5. Cancer Assessment Group Values (CAG)—These are recommended life-
   time exposure limits to known carcinogens which have been developed
   by the USEPA for a limited number of toxic compounds. The CAG
   number represents maximum allowable concentrations that may result
   in incremental  risk of human health over the short-term or long-term
   at an assumed risk. This assumed risk (10"') is the expected number of
   increased incidences of cancer in the effected population when the con-
   centration over a lifetime equals the specified  value. The CAG values
   are listed in the "Land Disposal Toxic Air Emissions Evaluation Guide-
   line" published by the USEPA in Dec. 1980.
6. This number related the TLV which assumes 8 hr/day, 40 hr/week ex-
   posure to 24 hr/day, every day exposure. In addition, it also includes a
   safety factor of 100.
7. Philadelphia Air Management Services, "Recommended Ambient  Air
   Quality Guidelines for Toxic Air Contaminants," June 1983.
 184
CONTAMINATED OROUNDWATER CONTROL

-------
        INPLACE  CLOSURE OF PREVIOUSLY BACKFILLED
                 AND  ACTIVE SURFACE  IMPOUNDMENTS
                                             WAYNE CRAWLEY
                                            K.W. BROWN, Ph.D.
                                             DAVID ANDERSON
                                      K.W. Brown and Associates,
                                            College Station, Texas
            Inc.
INTRODUCTION

  In-place closure should be evaluated as an option whenever other
cleanup options would involve removal of more than just a few
truckloads of waste and contaminated soil. Not only is removal ex-
pensive, but it also merely transfers the  hazardous material and
associated risk to a new  site. The in-place closure option can in-
clude treatment of the waste and other  contaminated media to
render them nonhazardous.
  In-place closure techniques can be used on impoundments that
have either reached the end of their useful life or have been inade-
quately closed. The case history discussed here involves both of
these applications for in-place closure:
•Two  older impoundments,  which  had  contained spent pickle
 liquor, had previously been inadequately closed by simply back-
 filling with soil (Fig. 1)
•Two large surface impoundments, also containing spent pickle li-
 quor, were to be taken out of service and closed by simply back-
 filling with soil (Fig. 1)
Separate investigative approaches and closure plans were developed
for each site. Following some background information, results of
the investigations and progress toward final closure of both the ac-
tive and previously backfilled impoundments are presented.


BACKGROUND
  Impounded waste at the site had been generated by the pickling
(scale and rust removal) of steel by dipping the  metal in a bath of
sulfuric acid. Typically, the acid bath was replaced when its iron
content (from dissolved rust) reached 9 to 10%. The spent acid was
simply pumped into the  impoundments.  The build up of solids
eventually filled the impoundments, requiring new impoundments
to be constructed.
  All impoundments discussed in this report were constructed by
simply bulldozing soil outward from what would be the middle of
the impoundment. Soil removed from the middle was used to form
the sidewalls  and dikes. Prior to enactment of  RCRA, impound-
ment closure was simply the reverse of impoundment construction,
i.e., the dikes were bulldozed back into the impoundment.
  Pre-RCRA closure regulations did not require  a low permeability
cap, a cover of topsoil or  a permanent vegetative cover. Backfilling
of the two  older impoundments which  were  filled with  acidic
sludges resulted in subsidence of the backfilled material and oozing
of the sludge to the surface. Subsequent rainfall  resulted in erosion
of the backfill and overland flow of the acidic materials. In addi-
tion groundwater quality in the area of both the backfilled and ac-
tive impoundments has been degraded.
  Both the backfilled and active impoundments are located in a
marine deposit consisting of clay. This soil contains small amounts
of silt, sand, gravel, sandstone and iron ore. Below 5 ft, the soil
contains many calcium carbonate concretions, masses of soft lime
and some gypsum. A typical profile of the soil near the backfilled
impoundments is described in Table 1.
  Standard  engineering interpretation of borings down to 7 ft
would have suggested that the site was suitable for constructing im-
poundments. Hindsight  would indicate that,  considering the
calcareous deposits, the site might fail to contain acidic wastes.
Consideration of the profile characteristics below the 7 ft depth
would also have indicated the potential for pollutant migration
(Table 1).
ASSESSMENT OF THE BACKFILLED
IMPOUNDMENTS
  Objectives  of the assessment associated  with the previously
backfilled impoundments included:
      MW-2
      O
      MW-3
       O
IMPOUNDMENT I
I PREVIOUSLY
 BACKFILLED!
            IMPOUNDMENT 2
            (PREVIOUSLY
             BACKFILLED)
                                      WATER AND
                                  SURFACE WATER
                                  FLOW
            IMPOUNDMENT 3
             (ACTIVE)
                                  GROUNDWATER AND
                                  SURFACE WATER
                                  FLOW
                                   MW-4
                                   O
     MW-5
     O
     MW-6
            IMPOUNDMENT
             (ACTIVE)
                         Figure 1
       Locations for Monitoring Wells Associated with Active
            and Previously Backfilled Impoundments

               IN SITU & ON-SITE TREATMENT      185

-------
                           Table 1
     Typical Profile of the Soils Adjacent to the Impoundments
Oeptu
(HI
D.u- l.i
1.5- 3.0

1.0- 5.0



5.0- 7.5




75-90
9.0-10.5
10.5-13.0

Liquid
Limit
bo
1)8

131



95




so
51


PUltlcl ly
indei
21
44

94



62





JB
--

i Pdlllng
HO. !aa
Sieve
It,
56

97



80





82
99

pH
0.0
1.2

).8



8.2





e.o
7.3

dim-
f Ulllon
GC
CH

CH



CH





CH
SH

Ururlpllon
Brown Cllyey Gravel
Un ind P.»d Cllr
Mlldl, Alklllne
lln f«ct


                                                         Soluble
                                                                                              Soil
                                                                                                      Soil
                                                                  M
                                                                       9.81
lolil
iull.lci
p»
*1
Cd
Cr
Fe
fb
i.n

-------
  Groundwater leaving the site (downgradient) had elevated con-
centrations of sulfates and total dissolved solids (TDS). Values ob-
 amed for pH  and chromium  were similar for monitoring well
water samples collected upgradient and downgradient  (Table 3).
                            Table 3
    Analysis of Groundwater Samples Taken During Site Assessment
Monitoring
Well No.
1
2
3
Position
of Well
upgradient
downgradient
downgradient

pH
6.1
6.1
6.1
S04 =
(mg/D
90
1800
2000
TDS*
(mg/l)
550
3050
3150
Cr
(mg/l)
^0.02
< 0.02
50.02
• Total Dissolved Solids

CLOSURE OF BACKFILLED
IMPOUNDMENTS

  The assessment revealed that soil within and below the backfill
had elevated sulfates and low pH values. These conditions extended
to a depth of 12 ft over an area of 1.25 acres for a total volume of
approximately 24,000 yd3 of affected soils and sludge. Attempting
to remove this volume of material for treatment or off-site disposal
would be extremely  expensive. Consequently, it was decided to
evaluate methods for in situ treatment.
  The most promising method found for in situ treatment of acidic
wastes was lime injection. Positive attributes of this method in-
cluded low cost, availability and the fact that injection equipment
could maneuver on the present backfill. Lime injection is primarily
used for soil stabilization in areas with expansive clays.5 The tech-
nique uses a hydraulic system to  force perforated steel tubes into
soil and a high pressure duplex mud pump to inject the slurry. For a
horizontal  spread of 2.5  to 3.5 ft, one  injection sequence is
necessary for each 25 ft2 of area.
  After it was decided that the appropriate closure technique for
the backfilled impoundments was lime injection, the following five
phases were delineated:
 •Determination of the lime required to bring the  backfilled ma-
 terial to a pH of 6.5
 •Subsurface injection of lime (calcium hydroxide slurry) into the
 backfilled impoundments
 •Surface application of calcium carbonate (CaCO3) to the affected
 soil surfaces
 •Revegetation of the area
 •Continuation of groundwater monitoring
  The closure plan involved  injection of lime to a depth of 12 ft
over 1.25 acres of backfill. The main objective of this step was to
reduce movement of sulfates  by reducing the sulfate concentration
in the soil solution. Soil channels clogged with lime would reduce
the rate at which migrating liquids would enter groundwater. In ad-
dition, the increased soil  pH and calcium concentration in these
channels would tend to  decrease sulfate concentration in solution.
The level of sulfates in groundwater should be substantially re-
duced over time. The exceptionally high sulfate  levels in the soil,
however, will prevent the sulfate concentration  in downgradient
water samples from returning to the background value of 375 mg/l.
  Lime slurry was injected into the ground at a  pressure of 50 to
200 lb/in.2 Slurry initially  flowed into the soil through open joints,
fissures, fractures  and other  channels. These interconnected veins
of lime served to both increase pH and decrease permeability. The
slurry was made by blowing powdered calcium oxide (CaO) into a
 16,000 gal tank. Two pounds of CaO were mixed with each  gallon
of water to form a calcium hydroxide [Ca(OH)2] slurry.
  Sulfates migrated to the groundwater through the same channels
through which the slurry would flow. Consequently, lime injection
would treat the zone considered most critical (i.e., the pores and
ped faces associated with these channels).
  The total lime requirement was calculated as follows:
•A sufficient  number  of soil cores were collected to be  repre-
 sentative of the entire volume of backfill, sludge and subsoil to
 be treated.
•These cores were composited, and studies were conducted on
 the composited sample to determine the amount of CaO needed
 to bring the backfilled area to a pH of 6.5.
The total lime required to bring the entire volume of material to pH
6.5 was calculated to be 172 tons. The Ca(OH)2 slurry was injected
at 5 ft intervals, using a small tractor equipped with 12 ft injection
rods mounted on the front (Fig. 4). The cost of this operation was
$0.06/ft3 of soil, or approximately $32,000.
                            Figure 4
             Tractor Equipped with Lime Injection Rods

  Slurry was injected until refusal, the point at which the slurry
broke through the surface. After one pass across the area, 86 tons
of CaO had been injected, or one half of the lime requirement.
Cores of soil were collected in the treated area 24 and 48 hr after in-
jection to evaluate the effectiveness of the  lime injection. These
samples were taken between injection points. Free calcium slurry
was visible in nearly all samples, and pH values ranged from 3.0 to
11. Two cores had pH  values between 6.0 and 8.5 throughout.
Samples taken 8  months after the injections, however,  had pH
values that were only slightly higher than before the injections and
no free calcium was visible. As indicated in Figure 5, little change in
pH can be expected until injection of more than 75% of the lime re-
quired for neutralization.
  Sulfates in the groundwater decreased from 2,000 mg/l to about
950 mg/l, and TDS dropped slightly in both downgradient wells
one month after  treatment.  Sulfate levels began to increase four
months after the injections (Fig. 6). The initial decrease in sulfates
was thought to be due to the small amount of lime that was injected
directly into the groundwater. Long-term  decrease of sulfates and
TDS can be expected after the other half of the lime requirement is
injected into the backfilled impoundments.
  Liming of the surface soil with CaCO3 and  revegetation of the af-
fected  areas  have already  been completed.  A combination of
                                          25
                           TONS CdCOj/AC FT
                             Figure 5
                  Titration Curve for Backfilled Soils
                                                                                  IN SITU & ON-SITE TREATMENT       187

-------
         2600


         2400


         2200


         2OOO


         I8OO


         I6OO


         MOO


         1200


         IOOO


         BOO


         600


         400


         20O
                          IICCHO

                             TOTAL DISSOLVED SOLIDS
                          	SULfATCS
         PRETflCATMENT  (468
                            MONTHS

                           Figure 6
   Total Dissolved Solids (TDS) and Sulfate Concentrations (ppm) in
   Groundwater Wells Near the Previously Backfilled Impoundments

ryegrass and bermuda grass was planted. Groundwater monitoring
will continue for several years after the completion of lime injec-
tion.

ASSESSMENT OF THE ACTIVE
IMPOUNDMENTS
  Objectives  of the active  impoundment  assessment were  as
follows:
•To determine the strength of the impounded acid
•To evaluate the effectiveness of various liming agents that could
be used to neutralize the acids
Objective  number one was  met by collecting, compositing and
neutralizing samples of the acidic water in impoundments 3 and 4.
In a laboratory pilot study, between 14 and 17.5 g of calcium car-
bonate were  required  to neutralize one liter  of  the  impounded
acids.
  Liming agents used  in the laboratory  study included lime kiln
dust, fresh cement kiln dust, stockpiled cement kiln dust, waste ce-
ment from an oilfield drilling operation, calcium hydroxide and
agricultural lime (CaCO3) The lime kiln dust had good neutraliza-
tion potential, but it could only be obtained in a dry powder which
had to be blown across the  impoundments.  Due to unfavorable
prevailing winds, blowing the lime kiln dust would cause an unac-
ceptable air pollution problem.
  After samples of the impounded water were neutralized, they
were evaluated for both volume of resulting solids and quality of
remaining liquids. Based on  both criteria, calcium hydroxide was
chosen. Use  of calcium hydroxide resulted in formation of a
relatively small quantity of suspended solids and a water  quality
high enough to be discharged into the city sewage system.

CLOSURE OF THE ACTIVE IMPOUNDMENTS

  Closure of the active impoundments consisted of five phases:
•Neutralization of the  liquid  and pumping to the city
•Neutralization and dewatering of the sludges
•Investigation of the soils beneath the impoundments
•Placement of a cover in accordance with regulations to close im-
 poundments as landfills
•Continuation of groundwater monitoring
  Impoundment number 3 was  treated first. A slurry tank was
positioned  next  to  the  dike, and  calcium  hydroxide slurry was
pumped into the impoundment. Two gas powered air compressors
were used to aerate the  impoundment and aid mixing. There was
enough pressure from the slurry pump to keep the injection line on
or near the surface of the liquid in the impoundment. The line
moved around considerably; by relocating it to different sections of
the impoundment after each tank of slurry, the lime slurry was in-
jected in all areas of the impoundment. The resulting liquid had a
pH of 6.7 and an iron content of <0.1 mg/1. By allowing for settl-
ing, the total suspended solids value was also low enough to permit
the liquids to simply be pumped into the city sewage system.
  Neutralization of impoundment  number 4  did not go quite as
smoothly. Because of a drainage ditch between impoundment 4 and
the road, the slurry tank was parked approximately 225 ft from the
impoundment. This distance resulted in a significant pressure drop
in the injection line  which then sank to the impoundment bottom.
Consequently,  more mixing action was required. After liquids in
impoundment 4 were neutralized, they were discharged to the city
sewage system.
  Calcium slurry that was deposited on the impoundment bottom
during the process of neutralizing the impounded liquids had pH
values as high as 11. The original  plan was to dewater these sludges
by pushing in the impoundment  sides with a bulldozer.  However,
this proved very difficult due to the volume and  weight of the
sludge. It was, therefore, necessary to pump most of the sludge into
a diked drying bed for dewatering. Sludge in the drying beds was
disced  several  times to  aid drying. This process also mixed the
sludge with the underlying soil. The pH values of the sludges after
drying ranged from 6.3 to 8.3. This  mixture contained 47% silt and
about 6% clay. Once dry, the sludges were pushed back into the im-
poundment as  fill.
  Soil  samples were taken  across both impoundments  (3 and 4)
once they were  backfilled  with  the neutralized sludge and soil.
These samples showed basically a 4 ft zone of soils with pH values
between 4.0 and 6.7. At 4 ft below the surface, the pH of the soils
decreased to between 2.1 and 2.5.  Low pH soils extended to a depth
of 9 ft.
  Calcium hydroxide |Ca(OH)J slurry injection was again chosen
as the method of treatment for these buried soils. This time the en-
tire lime  requirement was injected.  That is, 5 ft centers were used
with a second pass beginning with injection points in the middle of
four previous injection points. One hundred and eighteen tons of
slurry were injected into this 5 ft thick zone (19.6 tons of CaO/acre
foot; 34.9 tons of CaCO3/acre foot). This procedure also left a 1 to
2 in. layer of calcium hydroxide on the soils, which was worked into
the soil surface.
  Three  feet of  compacted clay  soil were placed over  impound-
ments 3 and 4. The clay was then covered with 1 ft of topsoil, and
the topsoil  was planted  with winter ryegrass and bermuda grass.
Groundwater monitoring will continue until all values stabilize or
return to background levels.
REFERENCES
1.  Grim.  R.E., Clay Mineralogy. 2nd ed. McGraw-Hill, New York, NY,
   1968.
2.  Malcolm, R.L., Leenheer,  J.A. and Weed,  S.B., "Dissolution of
   Aquifer Clay Minerals  During Deep-Well Disposal of Industrial  Or-
   ganic  Wastes," presented  at the International  Clay Conference.
   Mexico City, Mexico, July, 1975.
3.  Brown, K.W. and Anderson, D.C., Effects  of Organic Solvents on
   the Permeability of Clay Soils.  USEPA. Washington, D.C.,  EPA-
   600/2-83-016, 1983.
4.  Lindsay,  W.L., Chemical Equilibria in Soils, John Wiley and Sons,
   New York, NY, 1979.
5.  U.S. Dept. of Transportation, Handbook for Railroad Track Stabili-
   zation  Using Lime Slurry Pressure Injection,  USDOT, Federal  Rail-
   road Administration, FRA/ORD-77/30, 1977.
188
IN SITU & ON-SITE TREATMENT

-------
                        ORGANIC  SLUDGE STABILIZATION:
                                  AN OPTION  THAT WORKS

                                            JOHN M. RADEMACHER
                                              CHARLES R. HANSON
                                           Velsicol Chemical Corporation
                                                 Memphis, Tennessee
INTRODUCTION

  Over the past 30 years, repositories for the residues of the manu-
facturing industry have accumulated to such an extent that there
are more than 170,000 pits, ponds and lagoons containing indus-
trial waste reported throughout the United States. The types of
chemicals, qualities and their effects or  potential effects on the
environment are unknown. Two environmental laws have been
passed which form the basis for addressing these situations—they
are the Resource Conservation and Recovery Act and the Compre-
hensive  Environmental  Response,  Compensation  and  Liability
Act. Operating within the framework of these laws,  industry is
developing approaches for securing some of these problem sites.
One such approach is illustrated in the following case study.

Velsicol's Past Disposal Practices
  As a result of the manufacture of resins and chlorinated organic
chemicals over a period of approximately  30 years, Velsicol's Mar-
shall, Illinois plant generated an 8.5 acre, 17-million gallon pond
of organic waste. The 5/6 Pond, as it is referred to, once consisted
of two ponds with an adjacent "bone yard"  of old equipment,
drums, pallets, etc., and covered a total area of approximately 22
acres. During the period from 1935 to 1966, almost all of the plant
effluent containing wastes from the resin and  chlordane  units
passed through these ponds, and the second phase materials were
retained in the ponds as  a sludge. Prior to the promulgation data
of RCRA, the use of these ponds was terminated.
process for  stabilization of the organic  sludge. This  process is
unique for the following reasons:
•Solidification/Stabilization—The  process  is  successful  in  the
 presence of sludges containing over 40%, by weight, of organics
 and allows the binding materials used to produce a permanent set
 in the mixture.  The overall expansion of the stabilized mass is
 essentially nil.
•Construction Costs—The process  allows for a minimum of con-
 struction equipment.  No exotic mixers or other unusual equip-
 ment are required, and the sludge being stabilized in this instance
 does not need to be pumped out to a laydown pad for stabiliza-
 tion. The speed  of the set is remarkable,  and construction equip-
 ment can operate on the area within  a short period of time.  All
 these factors, including the cost of materials, have produced a
 cost/benefit ratio that is more competitive than other technolo-
 gies proposed by outside firms.
•Health and Hygiene—The process allows use of a construction
 technique which minimizes handling as well as worker exposure.
 The  technique uses in-place stabilization which dramatically re-
 duces the vaporization of any volatile organics during the mixing
 activity.
•Leachate Generation—Application of this technology has resulted
 in improvements of the quality of leachate generated by passing
 through a column of the stabilized sludge. This improvement is
 especially significant  with  the less volatile chlorinated hydro-
 carbons.
SOLIDIFICATION/STABILIZATION
  In an effort to secure the site, Velsicol began an extensive pro-
gram aimed at the stabilization of these residual sludges in 1980.
This research project included the investigation of various solidifi-
cation methods. It was hoped that these methods would become
part of a total remedial measure program.
  Because the classical approach to stabilization failed, attention
was shifted  to developing methods of bridging  the materials so
that a cap could be constructed over the sludge.  Concurrent with
the investigation  of this second  approach,  evaluations of the
sludge's potential  for stabilization were sought from several con-
sulting firms. The result—at least five proposals  with varying de-
grees of effectiveness and with costs  ranging from $.35 to $1.007
gal for materials and labor alone were received.
  After becoming dissatisfied with bridging the  sludge as  an
acceptable approach, Velsicol intensified its efforts at stabilization
through an independent laboratory and developed a proprietary
Process Details
  As developed, the process centers on the use of certain moder-
ately reactive materials being mixed with sludge with high organic
content. The materials initially absorb the organics, including the
oils and resin in the sludge, rendering the sludge a granular, clay-
like material.  A setting reaction takes place, and the stabilized
sludge stiffens but remains very workable. These resulting materials
have a permeability of 7 x 10"' to 3 x 10 ~7 cm/sec, a California
Bearing ratio of 10-14 and an expansion ratio of less  than 1:1.1.
Other characteristics include a wet density of 91.3-132.9 lb/ft3 and
a dry density of 60.0-109.2 lb/ft3.
  Because the 5/6 Pond is non-homogeneous, it, like many others,
does not have  consistent  physical  or chemical  characteristics
throughout. The pond's stratified layers are rich in  oils, both
organic and inorganic, solids and emulsions.  Due to  this incon-
sistency, the preparation of custom formulas for various sections of
the pond have been required, and based on the physical character-
                                                                               IN SITU & ON-SITE TREATMENT       189

-------
istics of the sludge, field modifications to the laboratory-derived
formula  have been successfully made during  the stabilization
process. In order to assure effective treatment, the final physical
properties after curing are determined on a  random test basis.
  The application is accomplished by using standard earthmoving
equipment and allows for mixing in place  to a depth of 10 ft. The
5/6 Pond was originally  divided into  four sections by earthen
levees; these levees were used as the initial working platform. The
mixing of the additives with the sludge takes place by course on
the periphery  of the pond and the dividing levees using standard
earthmoving equipment. Since  the stabilized sludge sets within
three days and is able to withstand heavy  equipment traffic at the
time, an extended work platform is formed.
Closure

  Once stabilized,  the 5/6  Pond will be  capped. The  proposed
cover system differs from the capping of a typical landfill. One of
the major differences is that most municipal refuse is compacted in
cells,  and these can be built up to more closely match the desired
final contour.  In the case of the 5/6 Pond,  the residue to be capped
is a fluid sludge lying in a nearly level plane.
  In order to create the desired slope, the solidified material is
shaped to the appropriate slope prior to the placement of the com-
pacted clay cap. The fact that the stabilized material can be shaped
is a very important attribute. This characteristic allows the cap to
be  of  uniform thickness and  will minimize the  use of  expensive
clay. The cap  will consist of three basic layers. The first layer will
consist of 3 ft of compacted clay and will  serve as a barrier to the
penetration of water. The second layer will be made up of 0.5 ft of
gravely sand to serve as a drainage layer. To prevent washout at the
toe of the drainage layer, typar and rip-rap will be used to stabilize
these areas.
  For protection against erosion, the third  layer will be 0.5 ft of top
soil cover to support vegetation.  To convey surface water in a sheet
                                                         flow pattern away from the site, the final cover will be contoured.
                                                         The quality of the stormwater run-off from the cap will allow it to
                                                         be discharged to the small creek adjacent to Velsicol's plant site.
                                                           The proposed monitoring system for the site will consist of eight
                                                         wells to be used to monitor groundwater in the drift as well as the
                                                         bedrock. Six downgradient wells will be placed in three locations on
                                                         the westerly  side of the site. Two wells will be placed in one loca-
                                                         tion upgradient on the easterly side.

                                                         CONCLUSIONS

                                                           A permit  for stabilizing the pond sludges at the Marshall  site
                                                         was approved by the Illinois Environmental Protection Agency,
                                                         and on-site inspections have been conducted during implementa-
                                                         tion by both the  State and the USEPA which was apprised of the
                                                         process at the beginning.  Additionally, specialists  in  hazardous
                                                         waste  handling  from the  National Enforcement Investigations
                                                         Center visited the site and reviewed the process.
                                                           This stabilization process is, of course, only one part of an over-
                                                         all remedial program undertaken by Velsicol. Although the process
                                                         does not result in the conversion of waste from hazardous to non-
                                                         hazardous, it is an effective demobilization technique.
                                                           In pursuit  of Velsicol's goal to secure the 5/6 Pond, the overall
                                                         knowledge gained by the company has produced a process which
                                                         lends itself to the economical solution of similar sludge ponds hav-
                                                         ing a broad range of composition. Velsicol's stabilization technol-
                                                         ogy is cost-effective—for the Marshall 5/6  Pond project, less than
                                                         S.30/gal—and results in 15% improvement in leachate quality with
                                                         respect to volatile organics, shows little measureable expansion in
                                                         volume as opposed to 50-100% for several other processes and pro-
                                                         duces workable material capable of being readily relocated or put
                                                         into an on-site  disposal system.  Based  upon  the evaluation of
                                                         samples from several other sites, it has been determined that a wide
                                                         variety of industrial organic sludges can be  stabilized  using this
                                                         technology.
190
IN SITU & ON-SITE TREATMENT

-------
        IN SITU VITRIFICATION—A  POTENTIAL  REMEDIAL
           ACTION TECHNIQUE FOR  HAZARDOUS WASTES
                                                V.F. FITZPATRICK
                                                     J.L.  BUELT
                                                      K.H. OMA
                                                C.L. TIMMERMAN
                                           Pacific Northwest Laboratory
                                               Richland, Washington
INTRODUCTION

  In situ vitrification  (ISV) is an innovative technology  being
developed as a potential method for stabilizing transuranic (TRU)
contaminated wastes  in place. Although  the process is  being
developed for TRU contaminated wastes, it is envisioned that the
process could also be applied to hazardous chemical wastes.
  In situ vitrification (ISV) is the conversion of contaminated soil
into a durable glass and crystalline wastes form through melting by
joule heating. The technology for in situ vitrification is based upon
electric  melter technology developed  at the Pacific Northwest
Laboratory (PNL)  for  the immobilization of high-level nuclear
waste.1 In situ vitrification was initially tested by  researchers at
PNL in August, 1980 (U.S. Patent 4,376,598).2 Since then, ISV has
grown from a concept to an emerging technology through a series
of 21 engineering-scale  (laboratory) tests and 7 pilot-scale (field)
tests. A large-scale system is currently being fabricated for testing.
The program has been sponsored by  the U.S. Department of
Energy's (DOE) Richland Operations Office for potential applica-
tion to Hanford TRU contaminated soil sites.
  The ISV development program is utilizing three sizes of vitrifica-
tion systems. The distinguishing characteristics of each system are
power level, electrode spacing and mass of block produced, as
shown below:
System
             Electrode   Vitrified Mass
Power (kW)   Spacing (m)  Per Setting (t)
Engineering
Pilot
Large
30
500
3750
.023-0.36
1.2
3.5-5.5
0.05-1.0
10
400-800
  Major advantages of in situ vitrification as a means of stabilizing
radioactive waste are:
•Safety in terms of minimizing worker and public exposure
•Long term durability of the waste form
•Cost effectiveness
•Applicability to different kinds of soils

PROCESS DESCRIPTION

  In situ vitrification is a process for stabilizing and immobilizing
contaminated soils. To begin the process,  shown in Figure 1,
graphite electrodes  are inserted vertically into the ground in a
square array. Graphite is placed on the surface of the soil between
the electrodes to form a conductive path,  and an electrical current
is  passed  between  the electrodes,  creating temperatures high
enough to melt the soil. The molten zone grows downward, encom-
passing the contaminated soil and producing a vitreous  mass. Con-
vective currents  distribute the contaminants uniformly within the
melt.  During the process,  gaseous  effluents emitted from the
molten mass are collected by a hood over the area  and routed
through a line to an off-gas treatment system. When power to the
system is  turned off, the molten volume begins to cool. The pro-
duct is a block of glass-like material  resembling natural obsidian.
Any subsidence can be covered with uncontaminated backfill to the
original grade level.
                                                   U     OFFGAS  HOOD

                                                          II	i
                         Figure 1
              In Situ Vitrification Process Sequence

  The principle of operation is joule heating, which occurs when an
electrical  current passes through  the molten medium. As this
molten mass grows, resistance decreases; to maintain the power
level high enough to continue melting the soil, the current must be
increased. This is accomplished by a transformer equipped with
multiple voltage taps. The multiple taps allow more efficient use of
the power system by maintaining the power factor (the relationship
between current and voltage) near maximum. The  process con-
tinues until the appropriate depth is reached. Melt depth is limited
as the heat losses from the melt approach the energy deliverable to
the molten soil by the electrodes.
  To contain off-gases that are released from the melting process,
an off-gas hood that is operated under a slight vacuum covers the
vitrification  zone. The hood also provides support  for the elec-
trodes. The off-gases are routed  from the hood to a treatment
system which scrubs and filters hazardous components.
  A more detailed description outlining the power system design
and the off-gas treatment system follows.

Power System Design

  The power system design is similar for all three scales of the ISV
program. A transformer connection converts three-phase alter-
nating current electrical power to two  single-phase loads. The
single-phase loads are connected to two of the electrodes, which are
arranged in a square pattern, creating a balanced electrical load on
the secondary. The even distribution of current within the molten
soil produces a vitrified product almost square in shape to minimize
overlap among  adjacent settings. Multiple voltage taps and  a
balanced  load allow  a near constant power  operation  which
shortens run time and thus minimizes cost.

Off-Gas Treatment System

  In both the pilot- and large-scale systems, the hood collects the
off-gas, provides a chamber for the combustion of released
pyrolyzed organics and supports the four electrodes embedded in
the soil. Much of the heat generated during the ISV  process  is
released to the off-gas stream. This heat is removed in the off-gas
treatment system, so that the temperature of the gas which exits
after treatment is close to ambient.
  There are three major kinds of treatment for the off-gas system
(Fig. 2). First, the gases are scrubbed in two stages with a quencher
and tandem nozzle scrubber. These scrubbers  remove particles
                                                                             IN SITU & ON-SITE TREATMENT       191

-------
down through  the  submicron  range.  Second, the water in the
saturated gas stream is removed by a vane separator and condenser
followed by another vane separator. Third, the off-gas is heated,
insuring an unsaturated gas stream at a temperature well above the
dewpoint, and then it is filtered with two banks of high efficiency
particulate air  (HEPA)  filters. Both  the  pilot-  and large-scale
systems are trailer mounted and therefore mobile.
                           Figure 2
           Schematic for the Large-Scale Off-Gas System

  The off-gas treatment system required  for 1SV application to
hazardous chemical wastes will probably be simpler than that re-
quired for radioactive  application. Specifically, the requirements
for dewatering and filtering the gas stream would probably not be
necessary.  In  some special applications, where one of the con-
taminants becomes a toxic gas, it may be necessary to add a special
treatment stage such as a charcoal bed.

PERFORMANCE ANALYSIS

  The ability of the waste form to retain the encapsulated or incor-
porated radionuclides (some with very long half-lives) is of prime
importance to the  usefulness of the ISV process.
  Vitrified soil  blocks have  been  analyzed to determine  their
chemical durabilities with  a series of tests  including 24-hr soxhlet
leach tests. The soxhlet leach rate for all elements was less than 1 x
IQ-'g/cmVday, an acceptable value. These rates were comparable
to those of Pyrex®  or  granite  and much less than  those of marble
or bottle glass (Fig. 3).
  A 28-day Materials Characterization Center test (MCC-1)' was
also conducted  on a  contaminated soil  sample that had been
vitrified in the laboratory at 1600 °C. The overall leach rate of the
vitrified soil is comparable to the PNL 76-68 waste glass developed
for high-level  nuclear  wastes.' The measured release rate  of Pu
from the vitrified soil was 2 x 10 ~7 g/cmVday. Higher vitrifica-
tion temperatures  like those experienced  in the field (-1700 to
2000 °C) are expected to lower the Pu leach rate.
  PYREX
  VTTRIRED
  HANFORD SOIL
  QRANITE
  BOTTLE GLASS
            0       1       2      3      4      6       (
                   SOXHLET CORROSION RATE, g/cm3 d « 10*

 * Pyrex is a registered trademark of Corning Glass Works, Corning, NY.

                          Figure 3
             Leach Resistances of Selected Materials
                                                          Another indication of the durability of the ISV waste form is
                                                        found in a study of the weathering of obsidian, a glass-like material
                                                        physically and chemically similar to the  ISV waste form.' In the
                                                        natural environment, obsidian has a hydration rate constant of 1 to
                                                        20 junVlOOO yr.*  A value of lO/tmVlOOO yr, assuming a linear
                                                        hydration rate, yields a conservative estimate of a I mm  hydrated
                                                        depth for the ISV waste form over a 10,000 yr  time span. Since
                                                        hydration is also  the  initial mechanism of weathering,  the  ISV
                                                        block is  expected  to  maintain  its integrity  at least through this
                                                        10,000 yr time period.
                                                          Another important factor to consider in the waste form evalua-
                                                        tion is the migration of the radionuclides once they are a part of the
                                                        molten waste form. In the pilot-scale field tests,  the radionuclides
                                                        did not move beyond the vitrified block. Furthermore, analysis of
                                                        the blocks from the tests revealed that the radionuciudes did not
                                                        concentrate in the block but, instead, were uniformly distributed.
                                                        These factors are very important considerations for application of
                                                        ISV to chemical wastes containing toxic or heavy metals.
                                                          Far term (10,000 yr) performance assessments have been made to
                                                        determine the effectiveness of selective vitrication for immobilizing
                                                        high TRU concentration zones at a reference waste site at Han ford.
                                                        Scenarios evaluated included inadvertent and  intentional human in-
                                                        trusion,  transients and permanent residents  in the vicinity of the
                                                        waste site. For these scenarios, the vitrified zone was covered by an
                                                        engineered barrier, and this combination was compared to  sites
                                                        with no remedial action and sites with just an engineered barrier.
                                                          Results of the analysis showed that the amount of radioactive
                                                        material available for human ingestion was reduced by up to 10' for
                                                        the site that was  selectively vitrified and had engineered barriers. It
                                                        was  concluded that vitrification cannot prevent  human intrusion
                                                        into old  or abandoned waste sites,  but it can moderate its conse-
                                                        quences.  The groundwater pathway was not considered for this
                                                        analysis because of the characteristics of the Hanford site. Insight
                                                        into the  long-term performance, when the groundwater  pathway
                                                        may be significant, can be obtained from the leach data presented
                                                        in the preceding  paragraphs.
                                                          Specific data on the leach rate of heavy metals are beyond the
                                                        current scope of the ISV program; however, the data for  radioac-
                                                        tive  contaminated soils indicate the potential for using ISV to
                                                        isolate toxic and heavy metals from the biosphere.
                                                          The release of elements from the soil to the off-gas stream during
                                                        processing was also studied. This partitioning is usually described
                                                        as the decontamination factor. The higher  the  decontamination
                                                        factor (the mass of an element in the soil divided by the  mass re-
                                                        leased to the off-gas treatment system), the smaller the amount of
                                                        an element that is released from the soil during processing.
                                                          Based  on results from the pilot-scale system, it is estimated  that
                                                        for the large-scale system, soil-to-off-gas-hood DFs for less volatile
                                                        elements  such as Pu,  Sr and U will be 1  x 103 to 1 x 10*. More
                                                        volatile elements such as Cs, Co and Te should have DFs of about 1
                                                        x  102. Low boiling heavy metals such as Pb and Cd should have
                                                        DFs about 10. (Additional data on heavy metals presented later in
                                                        the paper.)
                                                          Element retention increases with depth of burial and the presence
                                                        of a  cold cap and  decreases with the presence of gas generating
                                                        materials. Decontamination  factors the off-gas treatment system
                                                        (hood to  stack) are as  follows: for the semivolatiles (Cs,  Co  and
                                                        Te),  1 x  104 and for the less volatile nuclides Sr and Pu, 1 x  10s.
                                                        Therefore, the soil-to-stack  DFs are 1 x  106 for  the semivolatiles
                                                        and 1 x  10* to 1  x 109 for less volatile materials.  For particulates,
                                                        the DFs are about 1 x  10''.

                                                        PROCESS PARAMETERS

                                                          PNL staff studied nine soils from waste sites all over the United
                                                        States to determine how varying soil properties affect the vitrifica-
                                                        tion process. None of the normal variations  in properties such as
                                                        electrical  and thermal conductivities, fusion temperature, viscosity
                                                        and  chemical composition   significantly  impact   ISV operation.
                                                        While soil moisture increases the power requirements and run time
                                                        of the ISV process, it is not a barrier to its use, having only a small
192
IN SITU & ON-SITE TREATMENT

-------
effect on the attainable melt depth. Soil moisture is an economic
penalty proportional to the amount of heat required to evaporate
trie water.
  The effect of materials buried with the waste itself, particularly
those that are commonly found in waste sites, has been considered.
These  materials  include metals,  cements and  ceramics,  com-
bustibles and sealed containers. While there are some limitations to
the ISV process due to waste inclusions, they are not significant.
The most  significant consideration is sealed containers  housing
highly combustible organics. A large number of such containers
could potentially increase the flow rate requirements of the off-gas
system.
  The processing  capabilities of  the large-scale ISV  system are
depicted in Figure 4. The width per  setting ranges from 3.5 to 5.5
m, with attainable depths of 10 to  13 m. The depths are calculated
on a conservative basis using nominally high heat losses. Metals can
occupy 70% of the linear distance between electrodes with only a
10% decrease in voltage. This value represents process testing to
date rather than the limit for the system. The void volume of 4.3 m^
and combustible packages of 0.9 m*  reflect the capacity of the off-
gas treatment system. The solid combustible concentration of 3,200
kg/m/setting represents a situation that might be encountered in a
typical landfill disposal operation.  The combustible liquid concen-
tration of 4,800 kg/m/setting again reflects the capacity of the off-
gas treatment system. There is a design factor of two associated
with all of the void volume and combustible loading numbers. The
design factor will be verified by field testing the large-scale system
in FY 1985.
                                                 LARGE-SCALE
                                METALS . 70 LINEAR %
                                AND 5 WT%

                                LIQUID COMBUSTIBLE
                                CONCENTRATION

                                VOID VOLUMES

                                COMBUSTIBLE
                                PACKAGES

                                SOLID COMBUSTIBLE
                                CONCENTRATION
                                (WITH 30% SOIL)
4800 kg/m/SETTING


4.3 m3

0.9 m3


3200 kg/m/SETTING
                           Figure 4
               Large-Scale ISV System Capabilities

ECONOMIC ANALYSIS
   The cost of using ISV as an in-place stabilization technique has
been estimated by Oma et a/.7 The cost estimate includes expenses
from the following four categories: site preparation activities, an-
nual equipment charges, operational costs (labor) and consumable
supplies such as electrical power and molybdenum electrodes. Five
different configurations  were evaluated including variations in
operating manpower  levels,  power  source costs  and heat  loss
assumptions used by the mathematical model to predict processing
efficiency. The cost comparison for vitrifying to a depth of 5 m for
a  reference  contaminated zone configuration is given in Table 1.
The process efficiency for vitrifying to a greater depth and a dif-
ferent contaminated zone configuration is lower.
   Cost of electrical power and the amount  of soil moisture in the
area being  vitrified can  affect the  economics  of the process
significantly. The influence of these two parameters on cost is
shown in Figure 5. At low electrical rates (i.e., $0.029/kWh), power
costs account for only 20% of the total operational cost. However,
at $0.049 kWh and $0.0825 kWh, power costs account for 30% and
40% of the  total cost, respectively. The energy cost has a ceiling at
$0.0825/kWh;  above this electrical rate, a portable generator can
be leased and operated at an equivalent electrical rate of $0.0825
kHw 2 Soil moisture increases the operating cost of the process by
requiring more energy to vitrify  a given volume of contaminated
                   soil because the water in the soil must be evaporated. This adds to
                   the electrical energy costs and the time required to complete the
                   process which, in turn, increases the cost contribution from labor.

                                               Table 1
                           Cost Estimates for Five ISV Large-Scale Configurations
Number
1
2
3
4
5
Site
Hanford
Hanford
Hanford
Generic
Generic
Power
Local
Local
Local
Local
Portable
Heat Loss
High
Average
Average
Average
Average
Manpower
Level
Average
Average
Above Avg.
Average
Average
Total Cost
of Soil
Vitrified,
1982 S/m3
187
161
183
130
224
Total Cost
of Soil
Vitrified,
1982 $/ft3
5.30
4.60
5.20
5.10
6.30
EXPERIENCE WITH HAZARDOUS/
ORGANIC MATERIALS
  During process evaluation with the 21 engineering- and seven
pilot-scale tests,  various  hazardous, simulated  hazardous  and
organic materials have been added to the test area to determine
system performance. Some of these materials are Co, Mo, Sr, Cd,
Cs, Pb,  Ce,  La, Te and Nd as nitrates;  chlorides and oxides;
organic solvents such as carbon tetrachloride, tributyl phosphate
and dichlorobenzene; and combustibles such as cotton and rubber
gloves, wood chips and paper. The three main conclusions drawn
fromthese tests are: (1) burial depth attenuates release (e.g.,  1 to
IVi m of uncontaminated  overburden lowers release fractions
significantly);  (2) gaseous releases  associated with combustibles
result in a significantly higher release fraction; and (3) organics are
pyrolyzed, resulting in combustion in the hood directly above the
molten zone.
  The importance of burial depth during pilot- and engineering-
scale testing is illustrated in Figures 6 and 7, respectively.
  Gaseous releases enhance the release fraction. Once the material
is vitrified and incorporated into the  vitreous  mass, it is not
available for further release except in direct proportion to its vapor
pressure and in inverse proportion to its solubility in molten glass.
However, gaseous release, which is usually associated with combus-
tion, provides an additional release mechanism—entrainment—for
those  contaminants  associated  with  the  combustibles.  This
phenomenon can be seen by comparing Figure 7 with Figure 8 with
respect   to  Pb  and  Cd.  In these experiments,  the process
temperature was 1700 to 2000°C; thus,  for low boiling, insoluble
heavy metals, the release fraction can be up to several percent com-
pared to semi- and non-volatile elements. Again,  the release frac-
                          400
                          300
                          200
                          100
                                Utility ,
                                Power
I	^ Portable
     Generator
                                            468
                                            Electrical Rates C/kWh)
                                                                    10
                                                         12
                                              Figure 5
                           Cost of In Situ Vitrification as Functions of Electrical
                                       Rates and Soil Moisture
                                                                                   IN SITU & ON-SITE TREATMENT
                                                                             193

-------
tion will be dependent upon the vapor pressure and the solubility in
the glass.
  Combustibles testing has included up  to  50  kg of solid com-
bustibles and 23 kg of tributylphosphate in a single experiment.
Chromatographic, sample bomb and mass spectrometric analyses
of the effluent from both the hood and stack indicate less than 5 x
10-3 volume percent release  for  light hydrocarbon,  indicating
nearly complete pyrolysis and combustion.

CONCLUSIONS

  The following conclusions can be drawn from the evaluation of
ISV technology:
  * 999
  o
                           01     OB

                          BURIAL DEPTH 1ml
                           Figure 6
    Element Retention versus Burial Depth During Pilot Scale Tests
     g 999
     #996
                *     8     12     16     20     24

                  AVERAGE BURIAL DEPTH FROM SURFACE Icml
     100
  O   99


  J
  IU
  S   97

  i
  *   98
                   D
              4      8      12      18     20     24

                 AVERAGE BURIAL DEPTH FROM SURFACE (cml
                                                       28
                                                         •In situ  vitrification  is a developing technology that  may  have
                                                          significant potential  for selected hazardous waste disposal.
                                                         •Organic compounds are pyrolyzed during ISV. Subsequent com-
                                                          bustion and off-gas treatment  hold  potential  for  permanent
                                                          disposal of selected toxic organic wastes.
                                                         •Process economics for contaminated soil sites at Hanford are in
                                                          the range of $142 to S230/m3 ($4 to $6.50/ft3) of soil vitrified.
                                                          Differences in site geometry, electrical power costs, soil moisture
                                                          and other  factors can influence these  costs.  For a soil moisture
                                                          25%, and  using a portable  power supply, costs would be about
                                                          $320/m3 ($9/ft3) of soil vitrified.
                                                         •Long-term (10,000-yr) performance analysis  for TRU contam-
                                                          inants leads to the  belief that ISV may minimize the effects of
                                                          persistent toxic and/or heavy metal wastes.
                                                           When viewing the potential for ISV technology transfer from the
                                                         nuclear to  the hazardous waste arena, it is prudent to  remember
                                                         that ISV  appears to be an excellent specific remedial action techni-
                                                         que—it is not a panacea but, judiciously applied, the process holds
                                                         promise to mitigate the effects of unprocessed buried chemically
                                                         hazardous wastes.
                                                                            30
                          Figure 7
Element Retention versus Burial Depth During Engineering Scale Tests
                                                                                    20      30
                                                                                     RUN TIME h
                                                                                      Figure 8
                                                                      Cd and Pb Release as a Function of Run Time
                                                         ACKNOWLEDGEMENT

                                                           Work supported by the U.S. Department of Energy under Con-
                                                         tract DE-AC06-76RLO 1830.

                                                         REFERENCES

                                                         1. Buelt, J.L. el al., A Review of Continuous Ceramic-Lined Metiers
                                                            and Associated Experience at PNL. PNL-SA-7590, Pacific Northwest
                                                            Laboratory. Richland, WA, 1979.
                                                         2. Brouns. R.A., Buelt, J.L. and Bonner, W.F., "In Situ Vitrification
                                                            of Soil." U.S. Patent 4,376,598, 1983.
                                                         3. Materials  Characterization Center (MCC),  Nuclear Waste Materials
                                                            Handbook—Waste Form Test Methods. DOE/TIC-11400, Department
                                                            of Energy, Washington, D.C. 1981.
                                                         4. Ross, W.A. et at..  Comparative Leach Testing of Alternative TRU
                                                            Waste Forms. PNL-SA-9903, Pacific  Northwest  Laboratory, Rich-
                                                            land, WA. 1982.
                                                         5. Ewing,  R.C.  and Hoaker,  R.F., Naturally Occurring  Glasses: Ana-
                                                            logues for Radioactive  Waste Forms.  PNL-2776,  Pacific Northwest
                                                            Laboratory, Richland, WA, 1979.
                                                         6. Laursen, T. and Lanford, W.A.,  "Hydration of Obsidian."  Nature
                                                            276(9),  1978,  153-156.
                                                         7. Oma,  K.H.,  et  al., In Situ Vitrification of Transuranic  Wastes:
                                                            Systems Evaluation and Applications Assessment.  PNL-4800,  Pacific
                                                            Northwest Laboratory, Richland, WA,  1983.
194
IN SITU & ON-SITE TREATMENT

-------
TREATMENT, SOLIDIFICATION  AND ULTIMATE  DISPOSAL
 OF  HAZARDOUS  WASTE STREAMS IN SALT  FORMATIONS
                                             RAY FUNDERBURK
                                                   PB-KBB Inc.
                                                 Houston, Texas
INTRODUCTION

  In the mid-50s, the scientific community began to recognize the
unique characteristics of rock salt as a potential host medium for
the disposal of hazardous waste. In 1956, the National Research
Council  Committee of the National Academy of Sciences recom-
mended that salt be investigated as a geologic medium for the long-
term retention of liquid high-level radioactive waste.1 These early
requirements for the examination of the feasibility of storing high-
level radioactive waste in salt led to the famed "Project Salt Vault"
experiment  initiated  in  1965  by the  Oak  Ridge  National
Laboratory.2 The objective  of  the study was  to  confirm the
feasibility of disposal of high-level radioactive waste  in salt by
determining if there was a change in the stability of salt under the
influence of extreme heat and radiation.
  Spent nuclear fuel rods were emplaced in the Carey salt mine in
Lyons, Kansas at a depth of about 1000 ft. For two years, the salt
was exposed to radiation approaching 8 X108 rad and temperatures
reaching 390 °F. The  conclusion of the experiment was that there
were no adverse effects on the  salt formation—the bedded salt
structure had absorbed the intense heat and radiation without any
measurable deformation.
  A more exciting experiment in salt  occurred in 1965 when the
Atomic Energy Commission and the Department of Defense ex-
ploded a 5 kiloton nuclear  device in the Tatum  salt dome  in
southern Mississippi. The salt absorbed the heat, blast  and radia-
tion effects of the detonation without breaching. The only fractures
found in the salt were near the point of detonation and they were
only a few feet in length.3
  These early experiments  involving radioactive waste generated
scientific interest in salt as a retention medium for non-radioactive
hazardous waste. In  1975, the USEPA conducted a study of the
feasibility  of utilizing mined  openings  for disposal of  non-
radioactive hazardous waste.4 One geologic medium studied was
rock salt. The study conclusions showed that salt was  one of the
most desirable geologic hosts for hazardous waste retention due to
its unique physical characteristics.
  Currently, there are two hazardous waste disposal companies ap-
plying  for  USEPA  permits to  construct  and operate  non-
radioactive hazardous waste disposal facilities in domal salt. One is
in the state of Texas atop the Boling salt dome near Houston, the
other is in Louisiana on the Vinton salt dome near the town of the
same name.
SALT CHARACTERISTICS
  Although a mineral, halite or salt is unlike other minerals in  its
physical characteristics. Foremost, it becomes plastic  when sub-
jected to extreme temperatures, generally above 390 °F. When a
pressure exceeding 12,000 psi is exerted on  the mass, it becomes
viscous. However, these physical changes work to the benefit of
containment. At lesser temperatures and pressures, salt  still tends
to move and this movement is referred to as creep. As a result, a
salt mass can be fractured and, in time, it will close the fracture
along the fracture plane.  In other words, it is self-healing. This
characteristic makes salt ideal as a retention vessel for hazardous
waste. Compare this characteristic to hard  rock such as granite.
When hard rock is  fractured, it remains so and thus creates an
avenue for migrating fluids.
  Salt's lack of permeability is also attractive for retention of
hazardous  waste; a  permeability so low that it is considered im-
permeable. Because  it will not permit the passage of fluids, from
within  or  from outside  the  mass, it assures there will be  no
hydrological movement.
  Strength is another positive characteristic.  Salt averages 3000 psi
compressive strength which is equal to most construction concretes
in industrial use today.
  As a result, you have a deeply deposited geologic  medium that
is self-healing, is strong and will not permit the migration of fluids.
It is not surprising that the Federal Government is engaged in the
study and design of repositories  for high-level radioactive waste in
salt.

SALT DEPOSITS
  As illustrated in the accompanying map (Fig. 1), salt is found
throughout the United States in either bedded formations or in
massive salt domes.
  Bedded salt resulting from deposition of past ocean incursions is
the most common. The largest and thickest beds of salt  are found
in the Salina basin  encompassing the states of Michigan, Ohio,
New York, Pennsylvania and West Virginia. Beds of salt can be
found in Michigan, for instance, that are 1,800 ft thick. Further
west, underlying the states of Montana, North and South Dakota,
are salt deposits known as the Devonian salt basin. The Cimarron
and Permian salt basins are found stretching from Kansas through
Oklahoma down into Texas and New Mexico. Then there are the
Paradox and Supai basins in the far west.
  Of the occurrences  of salt, however, salt domes are the most
unique. Over 500 massive pillars of salt occur in the Gulf Coast
Embayment. About half are offshore stretching from southwestern
Alabama down to the  tip of Texas. They also extend further into
Mexico. These are huge masses of salt with diameters up to 10 miles
and depths exceeding 50,000 ft. It is postulated that these domes of
salt were created from pressures exerted on the mother salt laid
                                                                                      ULTIMATE DISPOSAL
                                                                                                                  195

-------
                                                          Figure t
                                               Salt Formations in the United States
down in huge quantities during the incursion of the Gulf of Mexico
which flooded most of the south and southwest in past geologic
time. As shown in Figure 2, these upward movements of salt resem-
ble what might take place if one were to quickly push downward a
hand in a pan of bread dough. Just as the dough would squeeze up-
ward through one's fingers, the  salt has moved upward due to the
imposed pressure.
  Salt domes have been the source of oil  and gas exploration for
many years. In fact, the first oil  gusher in Texas was discovered on
the periphery  of the famed Spindletop salt dome in Beaumont.
Because of the upward thrusting of the salt, geologic strata in the
adjacent area  are forced upward as well,  creating pockets of en-
trapment surrounding the  dome (Fig. 3). As a result of  the heavy
concentration of oil and gas exploration around  salt domes, much
is known of their character, size and  consistency.

CREATING VOIDS IN THE SALT

  There are two methods of creating a void in the salt to serve as a
retention vessel for hazardous waste. The first, and most common,
is physically mining the salt as is  done in a coal mine. However, this
procedure is labor  intensive and expensive in comparison with the
other method  called solution mining. In  solution mining, as the
term infers, one injects water  into the salt mass,  salt dissolves and
the resultant brine  is extracted thus creating a void. In actual prac-
tice, it is much more complicated. The following  is a description of
                                                        how a void—referred to as a salt cavern—is constructed by solution
                                                        mining.
                                                          A well is drilled into the salt formation to the prescribed depth.
                                                        Once the well is cemented into the strata  and salt, two suspended
                                                        casings are hung in the  open well, one inside the other (Fig. 4).
                                                        Fresh water is then injected into the inner handing casing and the
                                                        resultant brine is withdrawn through the annulus between the two
                                                        casings. This process continues until a sump has been created in the
                                                        salt cavern that  will eventually hold insolubles that are found in the
                                                        salt. The insoluble particles, mostly anhydrite,  fall to the bottom of
                                                        the cavern  by gravity and remain there throughout the develop-
                                                        mental process.
                                                          Once a sump has been created  of sufficient size  to  contain the
                                                        total insolubles to be released during creation of the entire cavern,
                                                        the hanging casings are  moved  to another position to further ex-
                                                        pand the walls  of the new cavern. At a certain stage,  the flow of
                                                        fluids is reversed with the fresh water being injected through the an-
                           Figure 2
            Schematic Diagram of Salt Pillar Formations
                                                                                    Figure 3
                                                                      Barbers Hill Dome, Mont Belview, Texas
196
ULTIMATE DISPOSAL

-------
nuius and the brine withdrawn up the inner casing. This process is
repeated until the cavern has reached the designed configuration.
  If the cavern is to be created in bedded salt, it will be horizontal
in shape due to the horizontal planing of the salt beds. Conversely,
if the cavern is constructed in domal salt, the vessel will be long and
tubular due to  the available depth of the salt. Once the cavern is
created, it is then voided of fluids and prepared for hazardous
waste disposal.
THE  DISPOSAL PROCESS
  In  the  family  of inorganic wastes,  there are basically seven
categories of chemically compatible waste. With the exception of
lithium and fluorides in heavy concentration, none of these groups
of waste react chemically with sodium chloride. As a result, they
can be exposed to  the salt surfaces of the  newly created cavern
without causing any undesirable chemical reaction.
  As  the inorganic wastes are received at the site, they are  batched
into chemically compatible groupings and held on the  surface until
sufficient quantities are on hand for processing.
  During the treatment process,  the waste stream is dewatered  if it
is too aqueous  and the resultant  moisture is retained for future use
in the slurry system to be described later. If the waste is acid,  it is
neutralized using a base,  hopefully utilizing another waste that is
basic. With the exception of neutralization and dewatering, very lit-
tle has to be done in the treatment phase to make the waste stream
ready for solidification.
  When sufficient volume of a particular group of waste warrants
processing and movement into a cavern, the disposal process is in-
itiated. First, the moisture content is determined and the fluid is
moved via pipeline to an automatic batch facility where a solidifica-
tion compound is added. Once the chemicals are mixed, the mass
becomes a slurry, is pumped to the wellhead over the cavern  and
moves by gravity and minimal pressure to the interior of the cavern.
  The slurried  mass must fall distances exceeding 2,000 to 3,000 ft.
As a result, there must be controls to prevent particulate  separation
during  the fall.  These  controls  are  manifested  in  finite
measurements  of the total fall  distance and manipulation of the
casing.
  As the slurry reaches the solidified layer of waste at the bottom
of the cavern that preceded the batch being pumped, it will tend to
seek  uniform layering and will fill any crevices or  rough edges
found in the salt wall. After the prescribed length of time, the mass
will harden to  a compressive strength averaging 3,000 psi  which is
equal to the compressive  strength of most salt formations.
  The time element involved in pumping batches into the same
cavern is in direct relation to the hardening  time consumed by the
preceding batch. In some cases, curing time will be as quick as 6 hr,
in others the time will be held to 48 hr due to the slurry consistency.
  When the cavern is almost filled, the remaining void will be used
as an anchor for the plug that will be created in the borehole (Fig.
5). The borehole,  with the cemented  casing  remaining, will be
pumped full of cement from the top of  the waste to within 10 ft of
the surface. Once the  cement plug hardens  and is tested for total
curing, the wellhead will be removed and  10 feet  of soil will be
emplaced over the  cement plug. At this point, the  cavern cycle is
complete and the surface can be returned to its original use.
   Such  a facility constructed over domal  salt would consume a
mere 20  acres of land yet  would be capable of disposing of an
estimated 25,000,000 gal/yr for 20 years. If the same surface facili-
ty were constructed over bedded salt, the caverns would be smaller
and the total land requirement would increase  proportionately.
  In  comparison with a landfill, one acre of salt disposal on the
surface equates to five acres of landfill 30 ft deep. Using the 5:1
ratio, it is obvious that land utilization is greatly enhanced by the
salt cavern alternative. In addition, the  land can be returned to its
original use when the salt repository is filled; that is not true for a
landfill.
COMPARISON  OF ALTERNATIVES
   There are no limitations if one uses the salt disposal technology
(Table 1). Salt  will  accept PCBs, dioxin,  heavy metals  and
     CAVERN DEVELOPMENT STAGES
                               PHASE
                                                  m
ISO  100  5O
 1111111111 11 III
            DEPTH
             (FT)
            -35OO
 50  100  ISO
111111 I 11 111|  -3600
                           -37OO


                           -J8OO


                           -39OO


                           -40OO


                           -4 ICO


                           -42OO


                           -43OO


                           -440O


                          -45OO


                          -4600


                          -4700


                          -4800
                                              MSO'
                                             o
                              4600'
                           CASING
                           CONFIGURATION
      CAVERN SHAPE
                        Figure 4
           Solution Mining to Form a Salt Cavern
                        Figure 5
        Filling of a Salt Cavern and Subsequent Closure
                                                                                               ULTIMATE DISPOSAL
                                                      197

-------
poisons—many  of  which  cannot  be  landfilled,  incinerated,
chemically processed, recycled or injected into a deep well disposal
system.

                            Table 1
           Hazardous Waste Disposal Option Limitations

Landfill
Incineration
Chemical
Processing
Deep Well
Injection
Salt Disposal
PCB
No
Yes
No
No
Yes
Dloxln
No
No
No
No
Yes
Heavy
Metals
No
No
Yes
No
Yes
Poisons
No
No
Yes
No
Yes
   From a cost standpoint (Table 2), the salt disposal system cannot
 compete with deep well injection which is a relatively cheap alter-
 native. However, there are some characteristics associated with
 deep well injection that minimize its use in certain instances. For
 example,  if  the  fluid to  be injected  into  a deep well  contains
 suspended solids 5 microns or more in size, the chances of suc-
 cessful injection are diminished considerably due to the plugging of
 the injection strata. Also, there is growing public and political con-
 cern over deep well injection as a continued alternative because of
 the possibilities of fresh water aquifer contamination.
                            Table 2
             Cost of Hazardous Waste Disposal Options
 Method of Disposal
 Landfill
 Incineration
 Chemical Processing
 Deep Well  Injection
 Salt Disposal
                               Cost ($/gal)
                               $0.20-$1.50
                                1.50- 8.00
                                3.00-15.00
                                0.15-0.50
                               $0.50 - $3.50
   The salt technology compares more favorably with the landfill
alternative in cost. However, it is capable of handling a larger vari-
ety of waste  streams  due to USEPA limitations on  substances
banned from being placed in a landfill.
   Salt would provide a natural repository for residue created dur-
ing incineration. And,  it would serve ideally as a system to support
chemical processing residue.
   "Disposal in salt is a technically attractive alternative,'" stated
Ronald D. Hill, Director of Solid and Hazardous Wasted Research
Division of USEPA's Cincinati group. Hill further stated that the
concept is technically  and economically  feasible and encouraged
the author to  continue research to bring  the technology to the in-
dustrial sector.
   Dr. Joel S.  Hirschhorn  of the Office of Technology Assessment
for Congress  stated  the  technology was "...several orders of
magnitude better than anything  on the horizon"  when  he  was
briefed on the concept.'
   USEPA Region VI, which controls the  states having most of the
salt domes in the Southwest,  is supportive of the technology for salt
dome disposal as another alternative. Many of the regulatory agen-
cies of states that have native salt are also supportive.
DOUBTS AND PERCEPTIONS
   When  presented  to  the lay person, this  concept envokes en-
thusiastic  support, but there are those that  challenge its  viability
with  respect  to either   destruction   of natural  resources  or
catastrophic failure due to  acts of God such  as  earthquake or
dissolution of  the salt mass from subsurface waters.
                                                           From a  natural resource standpoint,  there are sufficient salt
                                                         reserves to  sustain this nation for hundreds of years. In fact, there
                                                         is too much salt.  Most of the major salt companies that mine salt
                                                         for human  consumption are operating at much less than maximum
                                                         production levels. Many salt mines are standing idle due to the lack
                                                         of consumer demand. More importantly, the salt being considered
                                                         for hazardous waste disposal is too deep to be economically mined
                                                         for consumption. Most active salt mines are relatively close to the
                                                         surface, while salt deposits under study for disposal are from 3,000
                                                         to 5,000 ft  deep.
                                                           Many  have questioned  the  consequences of an  earthquake
                                                         should one  occur in the vicinity of a hazardous waste facility in salt.
                                                         Because of  the composition  of the salt mass—with its unique ability
                                                         to move and heal itself, it is  envisioned there would be little damage
                                                         within  the  formation  itself. One must remember that the energy
                                                         force of a 5 kiloton nuclear device is several orders of magnitude
                                                         more sudden and violent than the shifting forces of an earthquake.
                                                         Secondly,  because the hazardous waste  will  be  solidified  within
                                                         hours of being emplaced,  there will  be no opportunity for the
                                                         migration of hazardous fluids out of the formation even if it were
                                                         breached.
                                                           Lastly, there are those  who  understand the  process of salt
                                                         dissolution  from within to  create caverns, but fail to understand
                                                         why waters external to the salt formation will not dissolve the mass
                                                         and permit  exit of the waste.  For 200,000 to 300,000 years, these
                                                         salt masses  have been in contact with subsurface migrating waters.
                                                         These deep aquifers have come  in contact with the salt and have
                                                         become saline in the process. As a result, you find brine, but it has
                                                         become so saturated  with salt already that it  has very  little
                                                         detrimental effect on  the contact surface of  the  salt formation.
                                                         However, to ensure complete safety,  caverns are always created
                                                         with sufficient salt buffers between the cavern and the exterior of
                                                         the  formation so there will  be little chance of cavern  failure from
                                                         migrating waters.
THE FUTURE

  Certainly the future will see new processes developed to reduce
the amount of hazardous waste generated. The chemical industry is
meeting that challenge today—trying to develop compounds that
can be detoxified readily or used in another process so that hazar-
dous waste does not become a by-product, but rather  a product.
Unfortunately, it takes time to conduct research, and the nation's
time is running short.
  Incineration, both land and sea,  is costly but it offers almost
complete destruction. However, there are substances that will not
burn unless an inordinate amount of energy is expended to com-
plete destruction. And, there are substances that cannot be  in-
cinerated due  to  toxicity or their explosive  nature. Finally, even
with today's best incineration capabilities, there is residue that must
be placed somewhere.
  Two of the  cheapest forms of hazardous  waste  disposal—deep
well  injection  and landfill—are coming under growing political
scrutiny and their viability is being questioned. Many states have
already banned deep well injection of hazardous waste. Missouri,
for example,  forbids the emplacement of a hazardous fluid or
sludge  below  the surface.  Kansas recently passed a law that
precludes hazardous waste landfilling in the state. If this trend con-
tinues, and there is no reason to doubt that it will, the  number of
viable alternatives to disposal will shrink until chemical processing
and incineration will be the only near term alternatives  remaining.
Unfortunately, both  are quite expensive.
  Despite the exhaustive efforts of the USEPA to regulate landfill
operations and enforce compliance, many feel that current landfills
will become "Superfund" sites of the future. This  feeling was  ex-
hibited recently by Dr. Hirschhorn in a media interview where he
stated, "The presumption of many people is that you can regulate a
technology that does  not work in the first place. Land disposal does
not work."'
198
ULTIMATE DISPOSAL

-------
  Deep well injection continues to flourish, but the question arises
concerning the long term ability of the subsurface environment to
accept massive quantities of contaminated fluids. Even the largest
sponge reaches a point of saturation and will no longer accept addi-
tional  moisture.  And,  despite  the technical proof  that con-
taminated aquifers  are shielded from migrating fresh waters, one
must  be concerned  that  movements  within  the earth's  crust,
gradual or sudden,  might create paths of migration allowing these
fluids to join.
  States such as Michigan, with over 500 contaminated drinking
water wells—some  of which  are municipal water systems—have
traced the major source  of  contamination to  emplacement of
hazardous waste in surface lands.' This is very unfortunate since
the entire lower peninsula is underlain with massive beds of salt that
are ideal for the creation of caverns. More ironic is the fact that the
thickest salt beds are beneath the counties with  the highest concen-
trations of hazardous waste generation.
  It is time that the United States begins to examine another alter-
native, one  that is  environmentally safe, geologically secure and
certainly more cost effective—salt.
REFERENCES

1.  The Disposal of Radioactive Wastes on Land, National Academy of
   Sciences, National Research  Council  Committee Publication 519,
   April 1957.
2.  Empson, P.M., Bradshaw, R.L., Boegly,  W.J., Jr., McClain, W.C.,
   Parker, F.L., and  Schaffer, W.F., Jr., Project Salt Vault: Design and
   Operation, Oak Ridge National Laboratory, Oak Ridge, TN, 1971.
3.  Rawson, D., Randolph, P., Boardman, C., and Wheeler, V., "Post-
   explosion environment resulting from the Salmon experiment." Journal
   of Geophysics Research, 71, 1966.
4.  Stone, R.B., Aamodt, P.L., Engler, M.R., Madden, P., Evaluation of
   Hazardous Wastes Emplacement in Mined Openings, EPA 600/2-75-
   040, USEPA, Cincinnati, OH, 1975.
5.  Personal correspondence with the author,  1984.
6.  Personnel correspondence with the author, 1984.
7.  "Hazardous Waste," Compressed Air, 89, No . 6, June 1984.
8.  Assessment of Groundwater Contamination: Inventory of Sites. Michi-
   gan Department of Natural Resources. Lansing, MI, July 1982.
                                                                                                ULTIMATE DISPOSAL
                                                            199

-------
 DISPOSAL  OF SHOCK  SENSITIVE/EXPLOSIVE  CHEMICALS
UTILIZING EXPLOSIVE DETONATION WITH OPEN BURNING

                                             JOSEPH  S. BUTTICH
                                              SAMUEL J. GIANTI
                            New Jersey Department of Environmental Protection
                                   Hazardous Site Mitigation Administration
                                               Trenton, New Jersey
 INTRODUCTION
   In February 1979, prior to the availability of commercial facil-
 ities  to handle explosive and extremely reactive compounds, the
 New Jersey Department of Environmental Protection "Hazardous
 Chemical Explosives Team" developed a program to mitigate the
 statewide laboratory picric acid emergencies. At the request of New
 Jersey Emergency Response agencies, a federally permitted facility
 which frequently uses explosive detonation with open burning pro-
 cedures to eliminate spent materials agreed to assist the state in this
 disposal operation. After numerous detonation operations were
 performed, it was determined that this procedure would prove use-
 ful in the final disposal of  other materials  from the cleanup of
 hazardous sites.
   After  the original  picric acid  incident was  mitigated, the ex-
 plosives  team continued  to provide emergency assistance to insti-
 tutions across the state as a means of safely removing and dispos-
 ing potentially explosive/reactive laboratory chemicals such as
 organic peroxides, ethers and nitrated organics.
   This program has been in existence for Five years. The staff of
 the New Jersey Explosive Team has performed this operation num-
 erous times to dispose of thousands of pounds of explosive and re-
 active materials in a safe and environmentally sound manner.

 DESCRIPTION OF PROBLEM
   A  major misconception  associated  with  laboratory  chemical
 storage is that chemical  reagents last forever. Many laboratories
 have been left abandoned and many laboratory technicians have
 been injured because of  a cavalier or unknowing  attitude toward
 chemical storage. Laboratory storage has been and continues to be
 a common national problem.
   The problem develops in  the  following manner. Laboratories
 purchase large quantities of  chemicals to perform experiments or
 to develop new products. After  the projects are completed, the
 remaining chemicals sit on shelves in the laboratory to be forgotten
 until the time when personnel perform inventories and discover
 these materials several years  later and after their expiration dates.
 Additionally, local chemical companies  donate unwanted labora-
 tory  chemicals to area high schools which will never use these ma-
 terials. Many laboratory chemicals can be allowed to sit on shelves
 for many years without any substantial  change in chemical  prop-
 erties, but other materials such as oxidizers,  organic peroxides and
 ethers are not as  forgiving when it comes to age, contamination
 and  change  in  chemical  characteristics.   The  problem  when
 addressed properly and  efficiently is easily resolved but, if un-
 addressed, develops into an extremely hazardous situation.
                                                   DISPOSAL METHODS
                                                     Detonation with open burning is a relatively new procedure for
                                                   disposing of degraded potentially explosive and/or highly reactive
                                                   laboratory wastes. In the past, wastes have been disposed by pour-
                                                   ing chemicals down the drain,  illegal  dumping,  sanitary landfill
                                                   disposal or other suspect means.
                                                     Explosive detonation offers an  ultimate disposal method  for
                                                   these materials with no chance  of future explosion or reaction
                                                   problems. There are many ways in which detonation can be car-
                                                   ried out and many are dependent upon the size and shape of the
                                                   containers involved. The procedures below represent the methods
                                                   utilized in the authors' testing operations.


                                                   Detonation Utilizing Linear Shape Charge

                                                     The use of linear shape charge allows this procedure to be util-
                                                   ized on highly reactive, known-chemical, compressed gas cylinders.
                                                   When gas cylinders become old, their valves sometimes become in-
                                                   operable; conventional methods of disposal are useless for remedy-
                                                   ing this problem because  of an inability to release the cylinders'
                                                   contents.
                                                     The  detonation  disposal  process as applied to gas cylinders
                                                   works in  the following  manner. A section of linear shape charge
                                                   is secured lengthwise on the gas cylinder and tied inline with deton-
                                                   ation cord to two containers of  flammable materials, such as hep-
                                                   tane, acetone, etc. (Fig.  1).
                                                     When the detonation is initiated, the shape charge cuts open the
                                                   gas cylinder, and, simultaneously, the  detonation cord  ignites the
                                                   flammable materials. The fireball resulting from this explosion gen-
                                                   erates intense heat as all the combustible materials released are con-
                                                   sumed  in the flames. A main advantage in detonating these  ma-
                                                   terials is that the problem  of organic air contamination is minimal
                                                   because of the heat and turbulence generated by the explosion. This
                                                   process has been tested  numerous times and has proven very effec-
                                                   tive when used on cylinders containing known pyrophoric gases.

                                                   Detonation Utilizing Military Explosives

                                                     Military explosives are preferred over commercial explosives be-
                                                   cause they are appropriately 25"% stronger. Explosives such as
                                                   military dynamite,  composition-4  and others are utilized when
                                                   detonating spent reagents containerized in glass, metal or other
                                                   types of easily opened storage vessels. The process of  detonating
                                                   these materials is relatively simple. The materials which are to be
                                                   detonated are placed in a detonation pit and are arranged in a man-
                                                   ner to ensure thai all materials are consumed in the blast. Military
 200
ULTIMATE DISPOSAL

-------
                                     DETOMATION COUP
                                               FLAMMABLE MATERIALS
                              Figure 1
    Compressed Gas Cylinder Linear Shape Charge Detonation Diagram
                MILITARY  EXPLOSIVES
NOTE: All military explosives are tied into separate detonators for safety.
                              Figure 2
                Military Explosives Detonation Diagram
cause  of  their  extremely  low  self-accelerating  decomposition
temperature (SADT).
  Benzoyl peroxide reaches its SADT at  120°F and explodes at
176 °F.' The initial testing was performed using the detonation with
military explosives procedure. The results of these tests performed
without in-line flammable materials demonstrated a  high order
detonation  and the presence of  a cloud of dense white smoke.
When the same procedure was performed  along  with non-toxic
flammable materials, a high order detonation occurred again, but
the visible air  emissions were substantially  reduced and disposal
efficiency was greatly increased.
Inorganic Oxidizers

  Oxidizers such as potassium chlorate,  potassium perchlorate
and ammonium perchlorate produce oxygen and support combus-
tion when involved in chemical reactions. These materials are used
primarily  in the production of fireworks and explosives. Oxidizing
materials, when used in  these tests, were detonated simultaneously
with materials  such  as picric acid, ethyl ether, 2, 4,-dinitrophenyl
hydrazine and other potentially explosive compounds. Oxidizers
are primarily  utilized in these operations to aid in the complete
combustion of the hazardous materials which are being disposed.
The test results using Oxidizers produced large fireballs, great quan-
tities of heat and reduced signs of incomplete combustion. Overall,
oxidizer testing  has proven  useful in  supplying  the additional
needed oxygen to aid combustion of the other in-line chemicals.
                              Table 1
    Self-Accelerating Decomposition Temperatures of Organic Peroxides
explosives are placed around all sides of the chemical pile and an
additional charge is placed on top to ensure total consumption
(Fig. 2).
  When performing this operation, non-electric blasting methods
are utilized for initiating the charge. This allows personnel to place
several charges  in selected areas to guarantee initiation of the ex-
plosion. After the charges have been placed,  the igniter  pins are
pulled and the charges are activated. Detonations of this type often
do not require the use of additional flammable materials to support
burning since this procedure is most often performed on ethers and
other  potentially explosive  flammable  materials.  Overall,  this
method using military explosives has proven extremely effective as
the final disposal method for numerous types of non-toxic, non-
refractory potentially explosive chemical classes.
MATERIALS TESTED
  Detonation disposal has  been tested on several classes of de-
graded laboratory chemicals. The following summarizes the groups
tested:
•Organic peroxides
•Inorganic oxidizers
•Pyrophoric gases
•Flammable liquids

Organic Peroxides
  Commercially produced organic peroxides may be sold as solids,
liquids, pastes, granules or powders. Inhibitors are added to these
peroxides to reduce  their  flammability and  potential  explosive
power. Organic peroxides, as they are exposed to heat or shock,
degrade and become extremely unstable. This  instability is caused
by the volatilization or settling out of their respective reaction in-
hibitors. Peroxides were  tested (Table  1) for this procedure be-
        Acetyl Peroxide (25%)

           Ben:oyl Peromde

        t-Butyl Hydroperoxide

            t-Butyl Acetate

          t-Butyl Perbenzoate

          di-t-Butyl Peroxide

          t-Butyl Peroctoate

      t-Butyl Peroxyisobutyrate

        t-Butyl Peroxypivalate

           Caprylyl Peroxide

          Decanoyl Peroxide

    2.4-Dichlorobenzoyl Peroxide

       Cumene Hydroperoxide

           Dicumyl Peroxide

           Lauroyl Peroxide

   para-Menthane Hydroperoxide

             MEK Peroxide

         Pelargonyl Peroxide

          Propionyl Peroxide

   Dnsopropyl Peroxydicarbonate

       Succimc Acid Peroxide
                      0  25  50  75 100  125  150  175  200  22  250 275

                                  Temperature (Deg. F.)
Self-accelerating decomposition temperatures (SADT) of organic peroxides 1


                              ULTIMATE DISPOSAL       201

-------
Pyrophoric Gases

  The disposal of  pyrophoric  materials is a much  simpler pro-
cedure compared to other materials such as peroxides and oxidizers
because they ignite spontaneously on contact with air. Pyrophoric
materials observed during these tests have proven to be the least
difficult  to dispose, primarily for the following reason: during the
initial explosion, the charge forces the gases to escape the cylinder
at an extremely fast rate causing a high degree of turbulence above
the gas cylinder and aiding in the overall efficiency  of the burn.
The linear shape charge method of detonation/venting was utilized
on tests involving pyrophoric materials. The results of tests utiliz-
ing di-n-butyl lithium and  tri ethyl  aluminum have demonstrated
complete combustion and intense heat production. During actual
tests involving pyrophoric gases, there were no substantial visible
emissions to the atmosphere observed aside from initial formation
of metallic oxide intermediates which were later destroyed.
Flammable Liquids

   Flammable liquids, especially those  in the ether family, repre-
sent a special problem because ethers, unless they are  stored under
a nitrogen blanket, are  notorious  for forming  peroxides. These
peroxides are formed  as a  result of oxidation initiated by heat,
distillation, extended storage, degradation of inhibitors and ex-
posure to light. Ether peroxides  are extremely unstable. They have
exploded after being struck by a thrown stone.' Ethers,  in test
detonations with military explosives, have been shown to produce
large quantities of heat with minimal air emissions. The tests were
often performed using oxidizing materials to ensure complete com-
bustion. The disposal of flammable liquids utilizing this procedure
has been proven very effective.

CONCLUSIONS
   The program, which has been operative for the past five years,
has been proven safe and environmentally sound. Problems may
exist with incomplete combustion of some materials and the re-
lease of inorganic byproducts, but,  overall, this process proves to
be the best available solution to the explosive chemical problem at
this time.
                                                          An additional step being investigated by the team is the incorpo-
                                                        ration of additional materials in the explosion to add heat to en-
                                                        sure complete combustion. The program, however, was directed
                                                        toward detonation of materials which could be easily combusted.
                                                        The program is  presently pursuing a method for performing sim-
                                                        ilar  operations  within a  detonation chamber where testing and
                                                        treatment of off-gases can be performed.

                                                        GLOSSARY
                                                          Composition-4 (C-4):  Military  plastic  explosive consisting of
                                                        cyclonite (trimethylene trinitramin) and a plasticizer, which itself
                                                        may or may not be explosive.1
                                                          Linear Shape  Charge:  A plane-symmetrical hollow charge (cut-
                                                        ting charge) is an explosive charge with a hollow space, which acts
                                                        longitudinal in the plane of symmetry (roof-shaped).2
                                                           Detonation Cord: Explosive cord initiated by a blasting cap with
                                                        a detonation velocity of 21,000 ft/sec.
                                                          SADT (Self-Accelerating  Decomposition Temperature):  The
                                                        temperature at which  the decomposition of a material proceeds by
                                                        itself.1

                                                        ACKNOWLEDGEMENTS
                                                          The authors gratefully acknowledge the assistance and encour-
                                                        agement offered by several colleagues. In particular, they  appre-
                                                        ciate the assistance of the following: Scott A. Santora, Quality
                                                        Assurance Section, New Jersey Department of Environmental Pro-
                                                        tection,  Trenton, New Jersey; George R. Weiss, Hydro-Nuclear
                                                        Services, Medford,  New  Jersey; Emergency Services Section, New
                                                        Jersey State Police, Trenton,  New Jersey;  Ronald Decker, Bureau
                                                        of Alcohol, Tobacco and Firearms, U.S. Treasury Department.

                                                        REFERENCES
                                                        1.  Meidl, J.A., Explosives and Tevic Hazardous Materials. Glencoe Pub-
                                                          lishing, Encino, CA, 1970.
                                                        2. Meyer, R., Verlag Chemie, Deerfield Beach, FL, 1981.
                                                        3.  Meidl, J.A., "Flammable Hazardous Materials" Glencoe Publishing,
                                                           Encino, CA, 1978.
202
ULTIMATE DISPOSAL

-------
 INCINERATION  OF EXPLOSIVES  CONTAMINATED SOILS
                                            JOHN W. NOLAND
                                             Roy F. Weston, Inc.
                                         West Chester, Pennsylvania
                                              WAYNE E.  SISK
                                                     COR
                                                USATHAMA
                                         Edgewood Area,  Maryland
INTRODUCTION

  The U.S. Army Toxic and Hazardous Materials Agency (USA-
THAMA), located in the Edgewood area of Aberdeen Proving
Ground, Maryland, has  dual responsibility  for lethal chemical
demilitarization and installation restoration. It also serves as the
lead agency within  the U.S. Army Materiel Development and
Readiness Command (DARCOM) for pollution abatement and en-
vironmental control technology development.
  In this role, USATHAMA routinely  conducts generic research
and development (R&D) studies with wide application to current
U.S. Army environmental problems. The incineration of explosives
contaminated soils (IECS) project is an example of one of the many
successful R&D efforts USATHAMA has conducted  throughout
the years.

BACKGROUND

  Large quantities  of wastewater  are  generated  during the
manufacturing of explosives and propellants; the loading, assembly
and packing  of munitions; and demilitarization  and washout
operations. These wastewaters (referred to as "red water" or "pink
water" due to their characteristic color) contain varying concentra-
tions of explosives. Standard practice in the past has been to
dispose of these wastewaters in settling lagoons at various U.S.
Army installations. Although current practice provides for in-plant
treatment of these wastewaters, the inactive settling lagoons at
numerous U.  S.  Army installations are a source  of potential
groundwater contamination.
  USATHAMA  is currently evaluating a number  of potential
remedial action options for future implementation. One option has
emerged as the most promising in the near term (i.e., for installa-
tions requiring remedial action within the next five years). This op-
tion is excavation of the soils followed by thermal processing in a
rotary kiln incinerator. The U.S. Army routinely incinerates pure
explosives and propellants; however, previous to this project, this
technology was undemonstrated on explosives contaminated soils.

PROJECT OBJECTIVES

  The primary objective of these tests was to demonstrate the ef-
fectiveness of incineration as a decontamination method for ex-
plosives contaminated soils.
  The secondary objectives of the project were to:
•Develop a data base and appropriate correlations  for designing
 and predicting the  performance of the incinerator as a decon-
 tamination method
•Determine the fate  of the explosives and metals in the contam-
 inated soils during/after incineration
•Measure pollutant levels in the stack gas to determine the air pol-
 lution control devices that would be required  for future inciner-
 ators to incinerate explosives contaminated soils
Task
1 Incineration Equipment/
2 Son Characterization*/
3 Development ol Detailed
4 Environmental Permitting
S Evaluation of Materials
Handling Procedures
6 Incineration Testing


1982
SEP








OCT








NOV








DEC







1983
JAN







FEB







MAR







APR






MAY
U.S.
lEPAf




JUN
Army R
leview




JUL
•view





AUG






SEP
_•»


OCT
•

NOV


DEC


                                                       Figure 1
                                                   Project Schedule
                                                                                    ULTIMATE DISPOSAL
                                                    203

-------
                           Table 1
          Characteristics of Explosives Contaminated Soils
Description
Soil Matrix
Moisture Content
Ash Content (as received)
Explosives Content*
(dry basis)
— TNT
— RDX
— HMX
—Other
— Total Explosives
Heating Value (as received)
Soil Type "A"
Sand
12-26%
44-83%

9-41%
0.02%
Not detected
0.03%
9-41%
50-2.400 Btu/lb
Soil Type "B"
Clay
25-30%
54-66%

5-14%
3-10%
0.6-1.4%
0.06%
10-22%
600-1,200 Blu/lb
•See Table 2 for the molecular structures of the explosives.


PROJECT DESCRIPTION

  In  August  1982,  USATHAMA commissioned  the  consulting
firm of Roy F. Weston, Inc. to develop and implement a program
to demonstrate the effectiveness of rotary kiln incineration  in
decontaminating explosives contaminated soils. This program had
seven tasks:
•Incineration equipment/test site selection
•Soil characterization/reactivity testing
•Development of detailed test plan/safety plan
•Environmental permitting
•Evaluation of materials handling procedures
•Incineration testing
•Evaluation of results
                                         The IECS Project Schedule is shown in Figure 1. In the following
                                       sections of the paper, the authors have presented a summary of the
                                       results of the  work done.
                                       Incineration Equipment/Test Site Selection
                                         After  a  comprehensive  survey of rotary kiln manufacturers to
                                       determine  the availability of appropriately sized test units, Therm-
                                       All, Inc. of Peapack, New Jersey was selected as the incinerator
                                       subcontractor for the project. A major innovation of this project
                                       was the decision to use a "transportable" incinerator (i.e., equip-
                                       ment disassembled, loaded on trucks, shipped  to the test site and
                                       reassembled) as  opposed  to  a "mobile" incinerator  (i.e., truck
                                       mounted)  or shipment of  the contaminated soils to a commercial
                                       facility.
                                         The test  site selected was a United States Army installation in Illi-
                                       nois which provided the following advantages:
                                       •Remote location well isolated from populated areas
                                       •Proximity to contaminated soils
                                       •Well controlled access and security

                                       Soil Characterization/Reactivity Testing

                                         In order to maximize the usefulness of the results of the project,
                                       USATHAMA decided to test contaminated lagoon soils from two
                                       separate installations  with  widely varying characteristics (Table 1).
                                       The two installations selected provided ranges of soil characteristics
                                       typical of most other U.S. Army installations.
                                         The contaminated  lagoon soils are  hazardous because they ex-
                                       hibit the characteristic of reactivity (i.e., potential for detonation
                                       or explosion). Testing conducted at Allegany  Ballistics Laboratory
                                       (ABL) in Cumberland, Maryland confirmed  that the lagoon soils
                                       are reactive and that special precautions were required in develop-
                                       ing materials handling procedures and equipment design.
                                                             Table 2
                                                  Molecular Structure of Explosives
               CH,
   TNT
                             C.H.NiOi
               NO.
          2.4.6 Trinitrotoluene
                                                   TNB
                                                                            C«HjN,O.
                                                                                              2.6 ONT
                                                                                                                       C.H.N.O.
                                                                                                        2,6-Dmitrotoluene
                   NO>
        O.N   , - M
           v    \
            1
C.H.NiO.
            NOi
  1 ,3.5.7-Tetramtro-Ocuihydro- 1 .3.5.7- Tetracycloocuine
                                                                           C.H.N.O.
                                                                                              2,4 ONT
                                                                                                                        CrHiN,O4
                             1.3-Omitrobenzene
                                                                                 NO,
                                                                            24-Omitroioluene
               NOi
                            CiH.N.Oi
  ROX
                                                  NB
                                                                                                                       C'H'N'°'
     1.3.5-Tnnitro, Hexahydro-1,3,5-Tnazine
                                                            Nitrobenzene
                                                                                NOi
                                                                          Tetrarutromettiyianitme
                                                 2-Amlno
                                                                          C.H.N,O4
                                                         2-Ammo-4,6 Dlnllrololuene
204       ULTIMATE DISPOSAL

-------
                                                                                % Oj, COz, CO, HC, NO*
                                                                                Explosives, Particulates, Metals. SOi
                            % Explosives
                            % C,H.N,S.CI
                            % Moisture
                            % Volatiles
                            % Ash
                            % Mefa/s
                            Heating Value
                            Total Weight
                 Sediment -

                   Fuel —
                                                         % O4 CO,. CO. HC,
                                                    Explosives, Particulates, Metals, HCI
                    Air
                   Particulates - % Explosives
                             % C.H.N.S.CI
                             % Mefafc
                             Tbfa/ Weight
                             EP Toxicity
                                                                                            Key
                                                                                            Tl - Temperature Instrument
                                                                                            Fl - Flow Instrument
                                                                                            PI • Pressure Instrument
•     % explosives
     % C.H.N.S.CI
     % Metals
     Total Weight
     EP Toxicity Testing
                                                              Figure 2
                                                  Incineration Test Schematic Diagram
Development of Detailed Test Plan/Safety Plan

   To provide meaningful evaluation of the incineration test results,
a test plan was developed and certain key parameters were selected
to be controlled and held at various levels during the testing:
•Soil feed rate
•Temperature in the primary combustion chamber
•Temperature in the secondary combustion chamber
   The above parameters directly affected the economics of in-
cineration, i.e., how much can be burned; how quickly can it be
burned; and how much fuel is required?
   Other test variables were held constant to the extent possible.
Test variables that could not be held constant were measured dur-
ing the test as illustrated in the test plan schematic diagram (Fig. 2).
   From the  outset, USATHAMA assigned personnel safety the
highest priority for this project. In this regard, a detailed site plan
and safety submission  were developed and reviewed and approved
by the Department of  Defense Explosives Safety Board.

Environmental Permitting
  Recognizing the importance of Federal and state environmental
concerns, the IECS project was structured to be fully responsive to
the requirements of RCRA and  the Illinois  Air Pollution and
Hazardous Waste Management Regulations. As shown in the proj-
ect schedule, the environmental permitting was an extremely rigor-
ous and time consuming process.
Evaluation of Materials Handling  Procedures

  The primary objective of this task was to evaluate, design and
implement materials handling procedures that emphasized person-
nel and environmental safety. There were four major goals:
•Minimize personnel contact with  the lagoon soils
•Avoid confining the lagoon soils (which could lead to detonation)
•Avoid any initiating forces  (i.e., friction, heating under confine-
 ment, etc.)
•Contain any spills and minimize contamination  of clean areas
  The test plan was developed assuming the use of a screw con-
veyor to feed the contaminated soils into the incinerator. However,
subsequent soil reactivity testing at ABL led to cancellation of the
screw conveyor due to safety considerations. A soils handling pro-
tocol and a bucket feed system was designed specificallly for this
test program which met all  of the test objectives and safety re-
quirements. During the course of the test program, the feed system
(Fig. 3), cycled over 4,000 times without a single failure.

Incineration Testing

  The incineration testing commenced on Sept. 19, 1983. Nineteen
daily tests were completed in 20 consecutive calendar days with no
time lost due either to incineration or sampling equipment failure.
A summary of the test conditions for each of the 19 runs is given in
Table  3. Since explosive contaminated soils had never been in-
                          Figure 3
        Cutaway Sectional View of the Thermall Incinerator
                                                                                              ULTIMATE DISPOSAL
                                                          205

-------
                                                            Table 3
                                 Definition of Test Matrices and Summary of Controlled Process Variables
Test Run Tot Dale
i 9/19
3 9/21
15 10/4
2 9/20
5 9/23
8 9/27
4 9/22
10 9/29
14 10/3
12 10/1
1 9/26
19 10/8
17 10/6
13 10/2
16 10/5
6 9/24
9 9/28
1 1 9/30
18 10'7
Matrix No.
0-1
1-1
1-2
1-3
1-4
1-5
1-6
1-7
1-8
1-9
2-1
2-2
2-3
2-4
2-5
2-6
2-7
2-8
2-9
Soil Feed Rite
(Ib/hr)
500
300
350
400
300
350
400
300
350
400
300
350
400
300
350
400
300
350
400
Primary Kiln
Temptralurt ( °F)
800
1200
1200
1200
1400
1400
1400
1600
1600
1600
1200
1200
1200
1400
1400
1400
1600
1600
1600
Secondary Chamber
Temperature ( "ft
1400
1600
1600
1600
1800
1800
1800
2000
2000
2000
1600
1600
1600
1800
1800
1800
2000
2000
2000
Soil Type (A or B)
A
A
A
A
A
A
A
A
A
A
B
B
B
B
B
B
B
B
B
cinerated before, a preliminary test run (Test Run No. 1) was con-
ducted at the proposed maximum feed rate (500 Ib/hr) and pro-
posed minimum primary  kiln temperature (800 °F) to see  if ex-
plosives breakthrough would occur in the stack gas. No explosives
were detected in the stack gas; however, explosives were detected in
the kiln ash, fabric filter ash and in the  flue gas entering the secon-
dary chamber. Therefore, subsequent test runs were conducted at
lower feed rates and higher primary kiln temperatures to ensure
that all explosives would be destroyed.
  After the formal testing was completed on Oct. 8, 1983, an addi-
tional 25,000 Ib of lagoon soils were incinerated from Oct. 10 to 15
(64 actual hours of incinerating soils).  The objectives of burning
the additional lagoon soils were two-fold:
•To thermally treat all lagoon soils that  had been excavated but
 not required during the formal  testing
•To determine  the  operational characteristics of the incinerator
 system  under  a  long-term,  steady-state production mode  of
 operation

RESULTS

  The IECS test project was extremely successful:
•It demonstrated that a "transportable" incineration system could
                                                         be disassembled, transported approximately  1,000 miles, be re-
                                                         assembled and be fully operational within 2 weeks
                                                        •Nineteen days of formal testing were completed within 20 consec-
                                                         utive calendar days with no lost time due to equipment failure
                                                        •An additional 6 days of operation were performed at  steady-
                                                         state conditions  with  no down-time due to equipment failure or
                                                         malfunction
                                                        •An explosives destruction efficiency of greater than 99.99% in the
                                                         primary kiln ash
                                                        •An explosives destruction efficiency greater than 99.9999 + % in
                                                         the fabric filter ash

                                                        •No detectable explosives in the stack gas; therefore, an overall
                                                         destruction and  removal efficiency (DRE) of 100%
                                                        •Stack emissions were in compliance with  all Federal and state
                                                         regulations including: SO2, HC1, NO,, CO and particulates
                                                        •Ash residues were not hazardous using all RCRA criteria;  as a
                                                         result, an application was made to the Illinois EPA  and the ash
                                                         residues were land applied in an area adjacent to the  incineration
                                                         test site.
206
ULTIMATE DISPOSAL

-------
              AN OVERVIEW  OF "WHO  IS DOING  WHAT"
      IN  LABORATORY- AND BENCH-SCALE HAZARDOUS
                      WASTE INCINERATION RESEARCH

                                              C.C. LEE, Ph.D.
                                         GEORGE L. HUFFMAN
                                  U.S. Environmental Protection Agency
                                               Cincinnati, Ohio
INTRODUCTION

  It was estimated that 57 million tons of organic hazardous/toxic
wastes are generated annually in the United States, and 90% of
them are disposed of by environmentally unsound methods, thus
posing a serious threat to human health and the environment.' The
Federal government  responded to the critical hazardous waste
problem  with the enactment of the Resource Conservation and
Recovery  Act  (RCRA) in 1976 (Public  Law 94-580),  Toxic
Substance Control Act (TSCA) in 1976 (Public Law 94-469), and a
comprehensive Superfund program in 1980 (Public Law 96-510) to
assure the reliable management of hazardous/toxic waste disposal
operations and dump site cleanup. The enactment of these laws has
intensified research  into  the  thermal  destruction  of organic
chemical waste and this research has accumulated a large amount
of useful information.
  This paper synopsizes the past and current efforts in the area of
small-scale research that has come about due to the passage of
RCRA and TSCA. The purpose of this paper is to indicate "who is
doing what" in the areas of incineration of hazardous wastes in
terms of generating research information and in planning future
programs.
  The complete paper covers the following topics in great detail:
•Past Research Activities Including:
  •Non-flame Thermal Decomposition Research by the University
   of Dayton Research Institute (UDRI)
  •Non-flame Thermal Decomposition Research by the Union
   Carbide Corporation
  •Laminar Flame Combustion of Chlorinated Hydrocarbons by
   the Illinois Institute of Technology
  •Thermal Destruction  of Chlorophenol Residues by  Environ-
   ment-Canada
•Current Research Activities Including:
  •Hazardous Waste Incineration Engineering Analysis
  •EPA In-House Research
  •Investigation of Gas-Phase Thermal Decomposition Properties
   of Hazardous Organic Compounds by UDRI
  •The Incineration Characteristics of Selected Chlorinated Hy-
   drocarbons by the Louisiana  State University
  •Non-flame Waste  Decomposition of Hazardous Waste by the
   Midwest Research Institute
  •Heterogeneous Catalytic Oxidation of Model Chlorinated Hy-
   drocarbons by the Massachusetts Institute of Technology
The complete paper, "An Overview of 'Who is doing What' in Laboratory- and
Bench-Scale Hazardous Waste Incineration Research," can be obtained from the
author at USEPA, 26 West St. Clair St., Cincinnati, OH 45268.
  •Oxidation of Model Waste Components in Supercritical Water
   by the Massachusetts Institute of Technology
  •Molecular Beam Mass Spectroscopic  Study  of Chlorinated
   Hydrocarbon Flames by the Illinois Institute of Technology

RESEARCH NEEDS

  In 1980, USEPA published a comprehensive list of hazardous
wastes,2'3'4 One year later (1981), under  RCRA, USEPA  pro-
mulgated the  standards  for operating hazardous waste incin-
erators.5 The key requirements of the standards are:
•An incinerator must achieve a Destruction and Removal Effici-
 ency (DRE) of 99.99%  for each  Principal Organic Hazardous
 Constituent (POHC) designated for each waste  feed. Initially,
 USEPA suggested the use of compound incinerability and  con-
 centration for selecting POHCs. It further suggested  that  heat
 of combustion (AHC) be used  as the measure of compound in-
 cinerability in its guidance manual to permit writers.6
•An incinerator burning hazardous waste containing more  than
 0.5% chlorine must remove  99% of the hydrogen chloride from
 the exhaust gas.
•An incinerator burning hazardous waste must not emit particu-
 late matter exceeding 180 milligrams per day  standard cubic
 meter.
  A trial burn test(s) (or data equivalent to a trial burn) is required
to demonstrate the ability of a hazardous waste incinerator to com-
ply with the above performance standards.
  In 1979, under TSCA, USEPA promulgated the standards for
operating PCB (Polychlorinated  Biphenyl)  incinerators.  The
general requirements of the standards are:'
•Liquid PCBs—Maintenance  of the introduced liquids for a 2-sec
 dwell time at 1200°C (± 100°C)  and 3% excess oxygen in the
 stack gas; or
•Maintenance of the introduced liquids for a 1 [/2-sec dwell time at
 1600°C (± 100°C) and 2% excess oxygen in the stack gas.
•Combustion efficiency shall be at least  99.9% computed  as
 follows:
    Combustion efficiency = Cco2/(Cco2  + Ceo) x  100
where
    Cco2 = Concentration of carbon dioxide
    Ceo = Concentration of carbon monoxide

•Non-liquid PCBs—The mass of emissions from the incinerator
 shall be no greater than O.OOlg PCB/kg of the PCB introduced
 into the incinerator. (This is equivalent to 99.9999% DRE).
•Similar to the RCRA requirements, a trial burn test(s) is required
 to demonstrate that the incinerator meets the above standards.
                                                                                    ULTIMATE DISPOSAL
                                                                                                              207

-------
  Dealing with these problems requires full understanding of ther-
mal destruction phenomena. Only on the basis of this full under-
standing can incineration regulations be realistically developed, and
can incinerator  controls be properly designed. This is why major
research in incineration  is  needed.
  In conducting its incineration research program,  USEPA's In-
dustrial Environmental  Research Laboratory  in Cincinnati,  Ohio
has set the following major research goals.  All of its research ef-
forts are aimed at achieving these goals.
'Assessing Current  Capabilities of Destructors
  To assess the performance capabilities (Destruction and Removal
Efficiencies—DREs, etc.)  of existing hazardous  waste  thermal
destruction devices (incinerators,  kilns,  boilers,  etc.)  as  the
technical foundation  for  Agency  policies  and  regulations with
respect to thermal destruction as a hazardous waste disposal op-
tion.
•Defining POHCs and PICs
  To provide the necessary scientific basis: (1) for selecting the ap-
propriate Principal  Organic Hazardous Constituents (POHCs) for
specification in  permits; and (2) for delineating those conditions of
operation required to prevent the formation of hazardous Products
of Incomplete Combustion (PICs).
•Understanding the Thermal Destruction Process
  To  develop  an adequate understanding of thermal destruction
chemistry and the engineering of thermal processes so as to be able
to: (1) characterize and assess the performance of full-scale thermal
destruction devices  from a minimum set of evaluative tests; and (2)
extrapolate performance  information from  one  waste  type to
another or from one scale or type of equipment to another (e.g., so
that small-scale test burns can reliably be used in permitting deci-
sions).
•Monitoring Compliance
  To   define  easily  monitored  incinerator  facility  operating
parameters (e.g., CO/THC ratio, CO/CO2 radio, etc.) which cor-
relate with system performance (e.g., Destruction and Removal Ef-
ficiency, PIC formation,  etc.) so as  to  allow rapid, reliable and
economical determination by enforcement officials of compliance
with permit conditions and so as to define the  necessary preventive
and corrective actions to avoid uncontrolled excursions from per-
mit conditions.
•Innovative Technologies
  To  improve the cost-effectiveness and  broaden the applicability
of thermal processing as a hazardous  waste disposal option.

DISCLAIMER

  This paper has been  reviewed by the  Industrial Environmental
Research Laboratory,  U.S.  Environmental  Protection Agency,
Cincinnati, Ohio and approved for publication. Approval does not
signify that the contents necessarily  reflect the views and policies of
USEPA, nor does mention of trade names or commercial products
constitute endorsement  or recommendation  for use.

REFERENCES
  1.  Bonner, T., el al.. Engineering Handbook for Hazardous Waste In-
    cineration. Publication No. SW-889, September 1981. Prepared for
    the USEPA Office of Research and  Development by the Monsanto
    Research Corporation.
  2.  Environmental Protection Agency, "Hazardous Waste Management
    System: Identification and Listing of Hazardous Waste; Interim Final
    Rule and Proposed  Rule," Federal Register 45 (138) Part II, 47832-
    47836 (July  16, 1980).
  3.  Environmental Protection Agency, "Hazardous Waste Management
    System: Identification and Listing of Hazardous Waste, and Interim
    Status Standards  for Owners and Operators of Treatment, Storage,
    and Disposal  Facilities; Final, Interim, and  Proposed Regulations,"
    Federal Register 45 (212) Part XI, 72024-72041 (October 30, 1980).
  4.  Environmental Protection Agency, "Hazardous Waste Management
    System:  Identification and Listing of Hazardous Waste—Finalizing
    the Lists of Hazardous Wastes  (261.31 and 261.32)) and Proposal to
    Amend (261.32)," Federal Register 45 (220), Part VII, 74884-74894
    (November 12, 1980).
                                                          5. Federal Register, p. 7678, January 23, 1981.
                                                          6. Environmental Protection Agency, "Presentation of a Method for the
                                                            Selection of POHCs  in Accordance with the  RCRA  Interim  Final
                                                            Rule, Incinerator Standards,"  January  23, 1981, Office of  Solid
                                                            Waste.
                                                          7. Federal Register, p. 31551-52, May 31.  1979.

                                                          Garner, F.H., el al., "The Effect  of Certain Halogenated Methanes on
                                                          Premixed and Diffusion Flames," 6th Symposium  (Int'l) on Combustion,
                                                          p. 802, The Combustion  Institute, 1956.
                                                          Fletcher, E.A., el al.,  "Chlorine-Fluorine Flames," Combustion and
                                                          Flame, 12,  115(1968).
                                                          Simmons, R.F. and Wolfhard, H.G.,  Combustion & Flame, 1, 155 (1957).
                                                          Hall,  A.R.,  McCourbrey, J.C. and Wolfhard,  H.G., Combustion  &
                                                          Flame, 1, 53(1957).
                                                          Fenimore, C.P. and Martin, F.J., "Flammability of Polymers," Combus-
                                                          tion & Flame, 10, 135 (1966).
                                                          Bernard, J.A. and Honeyman, T.W.,  Proc. Roy. Soc. A, 279, 248 (1964).
                                                          Hoare, D.E., Walsh, A.D. and Li, Ting-Man. Eleventh Symposium (Int'l)
                                                          on Combustion, p. 879. The Combustion Institute: Pittsburgh (1967).
                                                          Hoare, D.E. and Li, Ting-Man, "The Combustion of Simple Ketones I,"
                                                          Combustion & Flame,  12, 136, 145  (1968).
                                                          Duvall, D.S.  and  Rubey,   W.A.,  Laboratory  Evaluation  of  High-
                                                          Temperature  Destruction  of  Kepone  and  Related  Pesticides,
                                                          EPA-600/2-76-299 (December 1976).
                                                          Rubey, W.A.,  Design Considerations  for a  Thermal Decomposition
                                                          Analytical System, EPA-600/2-80-098 (August 1980).
                                                          Duvall, D.S.  and  Rubey,   W.A.,  Laboratory  Evaluation  of  High-
                                                          Temperature Destruction of Polychlorinated Biphenyls and Related Com-
                                                          pounds, EPA-600/2-77-228 (December 1977).
                                                          Duvall, D.S.,  el al., "High Temperature Decomposition of Organic
                                                          Hazardous Waste," Proceedings of  the  Sixth Annual  Research  Sym-
                                                          posium: Treatment and Disposal of Hazardous Waste, USEPA, Municipal
                                                          Environmental Research Laboratory,  EPA-600/9-80-011 (March 1980).
                                                          Dellinger, B., Duvall, D.S., Hall, D.L.,  Rubey,  W.A. and Carnes, R.A.,
                                                          "Laboratory  Determinations  of  High-Temperature  Decomposition
                                                          Behavior of  Industrial Organic Materials," Presented at  75th Annual
                                                          Meeting of Air Pollution Control Association, New Orleans (June 1982).
                                                          Graham, J.L., et al., "Design and Evaluation of the Prototype Packaged
                                                          Thermal Reactor System," Draft Report to EPA, 1984.
                                                          Dellinger, B., el al., "Determination of the Thermal Decomposition Prop-
                                                          erties of 20 Selected Hazardous Organic Compounds," Draft Report to be
                                                          published.
                                                          Lee, K., Jahnes, H.J. and Macauley,  D.C., "Thermal Oxidation Kinetics
                                                          of Selected Organic Compounds," Proceedings of 71st Annual Meeting of
                                                          the Air Pollution Control Association, Houston, TX (June 1978).
                                                          Lee, K., Hansen, J.L. and Macauley, D.C., "Predictive Model of the
                                                          Time-Temperature  Requirements  for Thermal Destruction  of Dilute
                                                          Organic Vapors," Proceedings of 72nd Annual Meeting of Air Pollution
                                                          Control Association, Cincinnati, OH (June 1979).
                                                          Valeiras, H., Gupta, A.K. and Senkan, S.M., "Laminar Burning Velocities
                                                          of  Chlorinated  Hydrocarbon-Methane-Air Mixtures,"  Combustion
                                                          Science and Technology,  Vol. 36, p .  123 (1984).
                                                          Senkan, S., tl al.,  "On the Combustion of Chlorinated Hydrocarbons,
                                                          Part 1: Trichloroethytene," Vol. 35, p. 187-202, Combustion Science and
                                                          Technology (1983).
                                                          Senkan, S.M., Robinson, J.M. and  Gupta, A.K.,  "Sooting  Limits of
                                                          Chlorinated-Hydrocarbon-Methane-Air Pre-mied  Flames," Combustion
                                                          and  Flame, Vol. 49, p. 305 (1983).

                                                          Senkan, S.M., "On the Combustion of Chlorinated  Hydrocarbons, Part
                                                          II: Detailed Chemical Kinetic Modeling of Intermediate Zone of the Two-
                                                          Stage Trichloroethylene-Oxygen-Nitrogem  Flame," Combustion Science
                                                          and Technology, Vol. 38, p. 197 (1984).

                                                          Laboratory Scale Flame-Mode Hazardous  Waste Thermal Destruction
                                                          Research. USEPA. Draft Report to be published by NTIS.

                                                          Grumpier, E.P., Martin, E.J. and Vogel, G., "Best Engineering Judgment
                                                          for Permitting Hazardous Waste Incinerators," Presented at ASME/EPA
                                                          Hazardous Waste Incineration Conference,  Williamsburg, VA, May 27,
                                                          1981.
 208
ULTIMATE DISPOSAL

-------
Tsang, W.  and Shaub, N., "Chemical Processes in the Incineration of
Hazardous  Waste. National Bureau of Standards.  Paper presented to
American Chemical Society Symposium on Detoxification of Hazardous
Wastes, New York, NY (1981).
"Thermal Destruction  of Chlorophenol  Residues,"  Technical Services
Branch, Environmental Protection  Service,  Environment Canada.  July
1983.

Miller, D.,  et at., "Incinerability Characteristics  of Selected Chlorinated
Hydrocarbons," Presented at the Ninth Annual EPA Research Symposium
Land Disposal,  Incineration and Treatment of Hazardous Waste. Ft. Mit-
chell, KY, May 2-4, 1983.
Senser, D. and  Cundy, V., "The Incineration Characteristics of Selected
Chlorinated Methanes."  Presented  at the  22nd ASME/AIChE  Heat
Transfer Conference," Niagara Falls, NY, April 1984.
Manning,  M.P.,   "Heterogeneous  Catalytic  Oxidation  of  Model
Chlorinated Hydrocarbons," Presented at the Environmental Control Pro-
cess State-of-the-Art Seminar,  Cincinnati, OH, July 22-24, 1981.
Manning,  M.P., "Fluid Bed  Catalytic  Oxidation: An Underdeveloped
Hazardous Waste Disposal Technology," To be published in Hazardous
Waste Journal by Tufts University, Medford, MA.
Timberlake, S.H., Hong, G.T., Simson,  M. and Modell, M., "Supercrit-
ical Water Oxidation for Wastewater Treatment: Preliminary Study of
Urea Destruction." SAE Tech. Pap. Ser. number 820872 (1982).
Connolly,  J.F.,  "Solubility of Hydrocarbons in Water Near the  Critical
Solution Temperature." J. Chem. Eng. Data, 11, 13 (1966).
Martynova, O.I., "Solubility of Inorganic Compounds in Subcritical and
Supercritical Water" IN: Jones, D. de G., and  R.W. Staehle, Chairmen,
High Temperature, High Pressure Electrochemistry in Aqueous Solutions,
January 7-12,  1973, The University of Surrey, England, National Associa-
tion of Corrosion Engs., Houston, TX, 131 (1976).
Modell, M., "Reforming of Glucose and  Wood  at the Critical Conditions
of Water" ASME Paper 77-ENAs-2 (1977).
                                                                                                    ULTIMATE DISPOSAL
                                                              209

-------
  METHODS  OF DETERMINING RELATIVE  CONTAMINANT
         MOBILITIES  AND  MIGRATION  PATHWAYS USING
                             PHYSICAL-CHEMICAL DATA

                                                KARL L.  FORD
                                                PAUL GURBA
                                        Ecology and Environment,  Inc.
                                               Denver, Colorado
 NEED FOR PREDICTING
 MIGRATION POTENTIAL

   Under section 106 of CERCLA, the USEPA must provide an
 assessment  of the health  hazards presented by hazardous waste
 sites. The degree of rigor demanded for the assessment is dependent
 on the administrative and legal requirements of the USEPA.  The
 assessment is commonly termed an endangerment assessment.  The
 USEPA has provided recent guidance on how to conduct an en-
 dangerment assessment.1'2  This  guidance  emphasizes use  of
 physical, chemical and biological  data about the contaminant to
 assess its potential to migrate off-site.
   The endangerment assessment (EA) is intended to quantify the
 degree of exposure to human  and environmental receptors from
 off-site migration of contaminants, including data on the potential
 for  off-site migration and exposure of  receptors  from  con-
 taminants.  Since evidence of  off-site human exposure is often
 unavailable or difficult to measure, a demonstration of the poten-
 tial for migration is crucial to the success of the EA.
   In this paper, the authors describe two  methods presently being
 used by the Ecology and Environment,  Inc. Field Investigation
 Team to assess the migration potential of organic contaminants. A
 case history illustrating application of these methods is presented.

 METHODS OF ASSESSING
 MIGRATION POTENTIAL

 Mobility Index

   The first method designed for assessing migration potential is the
 calculation of a mobility index.1 This index produces a number that
 is proportional to the contaminant's probability of escaping its
 point of origin  and migrating through the air or water. Required
 data include only the molecular weight, water solubility (mg/l)  and
 vapor pressure (mm Hg). The mobility index (MI) is calculated as
 follows:
       _ log (Water Solubility x Vapor Pressure)

                             Koc
                                             (1)
where the Koc is the organic carbon partition coefficient. The Koc
is proportional to the tendency of a dilute aqueous solution of an
organic  compound  to  adsorb onto organic carbon. Means of
measuring this  property have  been  discussed  by several in-
vestigators. 4'5'6
  Published data about the Koc often are unavailable, and time
and resources may preclude analytical measurement of the Koc.
Fortunately, many investigators have demonstrated ways of predic-
ting the Koc from water solubility or the octanol-water partition
coefficient. The  regression equation used to predict the Koc for
most organic contaminants is from Kenaga:6
                                                    where S is the water solubility of the compound in mg/l. The reader
                                                    is cautioned that other regression equations may better predict the
                                                    Koc for a particular class of organic chemicals and is referred to a
                                                    discussion of this topic by Kenaga' or Lyman.'
                                                      Examining the mobility index  further,  one finds that vapor
                                                    pressure and water solubility are proportional to mobility. For ex-
                                                    ample,  substances with  high vapor pressure are  more likely to
                                                    volatilize and escape into the air while substances with high water
                                                    solubility are more likely to leach into groundwater or surface
                                                    waters leaving the site of origin. Substances with high Koc partition
                                                    coefficients are more likely to adsorb onto soil, aquifer materials or
                                                    sediment and are less likely to migrate from the site of origin. A log
                                                    function is used only to reduce the size of the numbers, hence some
                                                    contaminants of low water solubility  or vapor pressure will have
                                                    negative mobility indices.
                                                      A data-file has been created  containing physical-chemical data
                                                    and mobility indices for more than 100 organic priority pollutants
                                                    and pesticides.  Mobility indices in the data-file range from - 17.0
                                                    to  6.0.  Based on the distribution of these mobility indices, the
                                                    following guidance on relative mobilities is offered:
                                                            Relative Mobility
                                                                 Index

                                                                 >5.00
                                                              0.00 to 5.00
                                                              -5.00 to 0.00
                                                            -10.00 to -5.00
                                                                <- 10.00
                                         Mobility
                                        Descriptor

                                     extremely mobile
                                       very mobile
                                      slightly mobile
                                         immobile
                                      very immobile
      log Koc =  -0.55 log(S) +  3.64
                                             (2)
Equilibrium Compartment Model

  A second method is used to assess mobility in a slightly different
but similar way. An equilibrium compartment model is employed
to predict the environmental compartments (e.g., air, water, soil,
sediment and biota) most likely to accumulate  the contaminant.
This approach allows the user to construct a distribution profile of
the contaminant in the environment and to identify the most likely
pathways of migration and exposure. The model used is the Level I
fugacity model of MacKay.'
  The model is a simulation of the partitioning that would occur in
a closed system such as a closed aquarium. After introduction of a
measured amount of contaminant into a study area of known com-
partmental volumes,  the model employs partitioning theory to
predict the proportion  accumulated by each compartment at
equilibrium. The Koc describes the partitioning  between soil, sedi-
ment and water; the octanol-water partition  coefficient describes
the partitioning between the biota and water; and Henry's law
describes  the partitioning between  air and  water.  Data re-
quirements are identical to those required for the calculation of the
mobility index.
210
FATE OF CONTAMINANTS

-------
CASE HISTORY

  The example described here involves a landfill in Colorado that
had  a history of receiving large quantities of liquid hazardous
wastes prior to closure in  1980. It was placed in the National
Priorities List in  1982. Routine site monitoring disclosed extensive
groundwater contamination and established patterns of movement
of the contaminated plume over time. In fact, the profile of con-
taminants with respect to time and location has served as an oppor-
tunity to examine the validity of the methods described above.
  The physical-chemical data and mobility index for five of the ma-
jor contaminants present in the groundwater at the site are found in
Table 1. The distribution profiles predicted  for each of the en-
vironmental compartments as derived from the fugacity model are
found in Table 2. Most of the  contaminants discussed here are
volatile and semi-volatile solvent-type wastes, but the methods are
believed to be equally applicable to base-neutral priority pollutants
as well.

                            Table 1
        Physical and Chemical Properties and Mobility Indices
                  for Five Landfill Contaminants


Contaminant
1,1 dichloro-
ethane
1,1,1 trichloro-
ethane
benzene
toluene
tetrachloro-
ethylene10
Vapor
Pressure
(mm Hg)1

180

96
95.2
28.7

14
Water
Solubility
(mg/l)'

5500

2440
1800
535

175


Koc

38.0

60.3
70.8
138.0

257.0

Mobility
Index

4.41

3.59
3.38
2.05

0.98
(Superscripts 2 and 10 are references)
                            Table 2
         Distribution Profile for Five Landfill Contaminants
Contaminant

1,1 dichloro-
ethane
1,1,1 trichloro-
ethane
benzene
toluene
tetrachloro-
ethylene
% Air


0.9

6.6
2.4
1.3

3.8
% Water


98.6

91.3
96.4
97.0

91.6
%Soil


0.38

1.60
0.93
1.30

3.40
% Sedi-
ment

0.13

0.53
0.31
0.43

1.13
% Biota


0.005

0.01
0.007
0.028

0.038
   If one assumed (incorrectly) that all of the waste was deposited at
the same location at the same time, these methods should be able to
qualitatively predict the relative mobility of the contaminants. For
example, from these data, one would predict that the chlorinated
ethanes would appear earliest in the groundwater downgradient. As
the  number of available  adsorption sites  became exhausted,
benzene, toluene and tetrachloroethylene would increase.
   An examination of the actual monitoring data in four down-
gradient wells was conducted to attempt to verify this projection.
The distributions of the five contaminants in the wells with time are
shown in Figures 1-5.
   All of the contaminant concentrations  show a general trend of
decreasing concentration downgradient. Of greater interest is the
change  in concentration with time.  Both 1,1 dichloroethane and
1,1,1 trichloroethane show an increasing concentration downgra-
dient. Benzene and toluene, with lower mobility indices, do not ex-
hibit this trend  at this time. It is anticipated, however, that these
contaminants will eventually display the same trend  since they are
also predicted to migrate, although at a lower rate.
  Tetrachloroethylene does not fit  the  pattern as well. With  a
mobility index of 0.98, it should migrate even more slowly than
benzene  or toluene.  Like the chlorinated ethanes,  however,  a
noticeable  trend  of  increasing concentration downgradient with
              COHC. U9/1
                                                                                              A-1H            B-211
                                                                                              «ELL LOCATIONS OOWCRADIENT :
                                                                                              Figure 1
                                                                                Groundwater Concentrations with Time
                                                                                         1,1 Dichloroethane
                                                                                 COC. U9/1
                                                                                              *-IH
                                                                                              •ELL LOCATIONS OOVNGRADIENT 1

                                                                                              Figure 2
                                                                                Groundwater Concentrations with Time
                                                                                        1,1,1 Trichloroethane


                                                                                 coc. 09/1
                           A-IU
                            •ELL LOCATIONS DOVNCRAOIEMT
                           Figure 3
              Groundwater Concentrations with Time
                      Tetrachloroethylene
                                                                                         FATE OF CONTAMINANTS
                                                          211

-------
OCT. 1983
                                                                                              Table 3
                                                                                Chemical Concentrations at Landfill Site
                           t-llt            »-2l<
                           •QJ. LOCATIONS OOmCIUOIOIT
                           Figure 4
              Groundwater Concentrations with Time
                           Benzene
                            •FLL LOCATICNS
                           Figure 5
              Groundwater Concentrations with Time
                           Toluene

 time is evident. This difference may be due to factors other than its
 physical-chemical properties as discussed below. According to the
 relative mobility guidance suggested above, all of the contaminants
 would be classified as very mobile.
   Factors other than those entering into calculation of the mobility
 index are  likely  to have major effects on  the  migration of con-
 taminants. Seasonal fluctuations in groundwater, location and time
 of waste burial,  biodegradation and nature of cover and contami-
 nant practices are also very important in assessing mobility. In the
 case of 1,1,1 trichloroethane and tetrachloroethylene, it appears
 that most of these wastes were buried downgradient of well A-105.
   The second prediction method involved an estimation of the par-
 titioning behavior  of  the  contaminants.  However,  from  the
 predicted distribution,  one would project  that  all  of the con-
 taminants would accumulate largely in the water compartment,
 with small but perhaps measureable concentrations in the air com-
 partment for:  1,1,1 trichloroethane, tetrachloroethylene, benzene
 and toluene.
   The landfill data are not complete enough in all environmental
 compartments to completely assess  the  validity of  the fugacity
 model predictions of likely migration and exposure pathways, but
 qualitative evidence is available for support of the model's find-
 ings. As the previous discussion indicates, groundwaters at the site
 contain significant concentrations of the contaminants. Ground-
 water concentrations for well A-115, surface water concentrations
 from the January, 1984 sampling and air concentrations from 1981
 sampling for the five compounds, are shown in Table 3. The water
 concentrations are considerably larger than the  air concentrations
 as is suggested by the partitioning of the contaminants in Table 3.
Contaminant
1,1 dichloroethane
1,1,1 trichloroelhane
letrachloroethylene
benzene
toluene
ND meant not detected.
Well A-105
6300
ND
ND
3600
6900

Sarhce Wiler
92
720
46
ND
ND

Air (DMT A-105)
ND
ND
ND
ND
1.5

                                                         Concentrations given in pg/1


                                                           Both air and water were anticipated to be important pathways
                                                         from the application of the compartment model. This information
                                                         along with the toxicity, carcinogenicity and other data helped
                                                         establish  the level of protection needed for on-site work. More ad-
                                                         vanced fugacity modeling is presently being applied to quantitative-
                                                         ly predict actual  compartmental  concentrations. Preliminary data
                                                         show good agreement between predicted and actual environmental
                                                         concentrations."

                                                         CONCLUSIONS
                                                           Two methods are described in  order to characterize the mobility
                                                         of organic  compounds using physical-chemical  data. One method
                                                         calculates a mobility index, describing the contaminant's mobility
                                                         in  the environment relative to other  well-known  or site con-
                                                         taminants. The second method uses an equilibrium compartmental
                                                         model to construct a profile of the distribution of the contaminant
                                                         in the environment.
                                                           Data from a NPL site in Colorado were used to evaluate the
                                                         suitability of these methods in characterizing environmental mobili-
                                                         ty. Groundwater monitoring of five contaminants downgradient as
                                                         a function of time has shown that contaminants with high mobility
                                                         indices move  more rapidly in groundwater than  contaminants with
                                                         lower mobility indices. Limited air and water monitoring support
                                                         the compartment  model predictions that these media are important
                                                         pathways of migration and potential exposure for the contaminants
                                                         studied.

                                                         REFERENCES
                                                          I. Morgan, R.C., Clemens, R., Davis, B.D., Evans, T.T., LiVolsi, J.A.,
                                                            Mittleman, A.L., Murphy, J.R.,  Parker, J.C. and Partymiller, K.,
                                                            Endangerment  Assessment  for Superfund Enforcement Actions,
                                                            1984.
                                                          2. Thomas, L.M., Endangerment Assessment Guidance, Memorandum
                                                            to Regional Administrators, USEPA,  1984.
                                                          3. Laskowski, D.A., Goring, C.A., Mcall, P.J. and Swann, R.L., "Ter-
                                                            restrial Environment in Environmental Risk Analysis for Chemicals,"
                                                            Environmental Risk Analysis for Chemicals,  R.A. Conway, ed., Van
                                                            Nostrand Reinhold Co., New York, NY, 1983, 198-240.
                                                         (''oKarichkoff, S.W. and Brown, D.S., Determination of Octanol- Water
                                                         ^^Distribution,  Water Solubility and Sediment-Water Partition Coef-
                                                            ficients for Hydrophobic Organic  Pollutants. USEPA.
                                                          5. Chiou, C.T., "Partition coefficients and bioaccumulation of selected
                                                          — ^organic chemicals." J. Environ. Sci. Technol., II, 1977, 475-478.
                                                                    E.E. and Goring, C.A., "Relationship Between Water Solu-
                                                             bility, Soil Sorption, Octanol-Water Partitioning and Concentration
                                                             of Chemicals in the Biota," Aquatic Toxicology. ASTM, Eaton, J.G.,
                                                             Parrish, P.R., Hendrichs, A.C., eds., 1980. 78-115.
                                                         vT)Lyman, W.J., Reehl, W. F. and  Rosenblatt, D.H., Handbook of
                                                             Chemical Property Estimation Methods: Environmental Behavior of
                                                             Organic Compounds, McGraw-Hill, New York, NY.
                                                          8.  MacKay, D., "Calculating fugacity." J. Environ. Sci. Technol., 15,
                                                             1981, 1006-1014.
                                                          9.  Ford, K.L., unpublished manuscript.
                                                         10.  Versar,  Inc.,  Water Related Fate of 129 Priority Pollutants, Vols. 1
                                                             and 11., prepared for USEPA, 1979.
212
FATE OF CONTAMINANTS

-------
        ENDANGERMENT  ASSESSMENTS  FOR SUPERFUND
                                 ENFORCEMENT ACTIONS

                                          R. CHARLES MORGAN
                                             ROBERT CLEMENS
                                            BARBARA D. DAVIS
                                             THOMAS T. EVANS
                                          JERALD A. FAGLIANO
                                          JOSEPH A. LiVOLSI, JR.
                                        ABRAHAM L. MITTELMAN
                                              J.  ROY MURPHY
                                              JEAN C. PARKER
                                         KENNETH PARTYMILLER

                                   U.S. Environmental Protection Agency
                                   Office of Waste Programs Enforcement
                                               Washington, D.C.
INTRODUCTION

  The  Comprehensive Environmental Response, Compensation
and Liability Act of 1980 authorizes the Federal Government to
respond directly to releases, or threatened releases, of hazardous
substances that may endanger public health, welfare or the environ-
ment (Public Law 96-510, 1980). Section 106(a) of CERCLA em-
powers the USEPA to compel responsible parties to clean up haz-
ardous conditions  through  administrative or civil order. If  re-
sponsible parties are not identified, the  USEPA may clean a site
using CERCLA "Superfund" money and seek recovery of costs
when and if responsible parties are later found.
  When undertaking cleanup or enforcement action at a given site,
the USEPA must demonstrate whether a potentially harmful situa-
tion exists and, if so, the kind and degree of endangerment. Sec-
tions 300.64  through 300.69 of the National Contingency Plan
(NCP) outline the factors that the USEPA must consider in assess-
ing this endangerment.5
  A USEPA endangerment assessment documents the adverse im-
pacts that could occur given potential or actual release of haz-
ardous material from a site. The assessment is  multidisciplinary
and may involve expertise from many technical areas such as chem-
istry, toxicology, geology, engineering,  environmental modeling,
demography and epidemiology.
  The level of detail of the assessment depends on the kind and de-
gree of perceived endangerment. The immediacy of risk in an emer-
gency situation may simplify the endangerment assessment when
the primary consideration is prompt  mitigation of a hazard or
potential for harm. When conditions allow time for more thorough
evaluation, a complete endangerment assessment can be performed
to better define the risk of harm and to consider options to
respond to the entire problem. Information obtained through this
process may be used to justify the choice of response action during
litigation for cost recovery.
  In this paper, the authors discuss the factors that should be con-
sidered in a complete endangerment assessment.

ENDANGERMENT ASSESSMENT FACTORS
  A complete endangerment assessment involves the gathering and
integration of the following information.
•Site history and management practices
•Identify, concentration and amount of hazardous substances orig-
 inally on the site
•Transport and environmental fate of hazardous substances on and
 off the site
•Toxicological properties of the hazardous substances
•Exposure to the human population and environment

Site History and Management Practices
  One of the, first steps in an endangerment assessment is defining
the exact location and areal extent of the site. Past and present pro-
duction and disposal uses of the site should be detailed, as well as
descriptions of process and containment facilities. Such manage-
ment practices as process conditions,  disposal methods and com-
pleteness of records should also be examined.
  An assessment should contain descriptions of topographic fea-
tures and locations of wells,  buildings, roads and water courses.
This information can be obtained from site visits and USGS quad-
rangle maps.
Identity of the Contaminants

  It is most helpful to determine, as precisely as possible, the spe-
cific chemical contaminants that are present at a site. Such infor-
mation may be  available  from company or government records;
however, some or all of this information is often lacking.
  The physical nature of contaminants known to be present should
also be determined (i.e., phase, oxidation state or whether mixed
or dissolved with other materials). This information may be useful
for predicting the fate of the contaminants, potential exposure
pathways and lexicological properties.
  The amounts of hazardous substances at the site should be esti-
mated using records of past activities. In addition, concentrations
of contaminants at or  near a site, in containers and in environ-
mental media, should be analyzed. Both factors, amount and con-
centration, are important in defining the overall endangerment.
  Samples for analysis should be taken to adequately represent the
degree and extent of contamination and should be reasonable and
appropriate for the situation. Comparable areas thought to be non-
contaminated should also be sampled  to give background  or
ambient concentrations for purposes  of comparison. It is impor-
tant that sampling and analysis are subject to strict quality control
procedures to ensure the accuracy of identification and measure-
ment.

Environmental Fate of Contaminants

  An endangerment assessment of hazardous substances at a waste
site requires an understanding of the likely movement, persistence
and transformation of the chemicals. Predicting the fate of haz-
ardous substances is an important step in estimating the potential
exposures to humans and the environment.  This prediction can be
                                                                                FATE OF CONTAMINANTS
                                                    213

-------
derived from knowledge of the geology, hydrology and meteorol-
ogy at a waste site as well as from the intrinsic properties of the haz-
ardous substances.1
  The  physical and chemical characteristics of a hazardous sub-
stance  affect its potential to be transported through air, soil, or
water or transformed into other substances. These characteristics
include molecular weight, vapor  pressure, melting and  boiling
temperatures, water solubility,  density,  Henry's Law constant,
octanol/water partition coefficient, adsorption coefficients, bio-
logical reactivity, absorption of visible and ultraviolet light, par-
ticle size distribution and dissociation constant.7
  Persistence of a chemical in a specific environmental medium is
an important consideration  in assessing potential exposures. Per-
sistence is a function of  the degree to which it is susceptible to
transformation processes  including aerobic and  anaerobic biode-
gradation, chemical reaction, complex formation, hydrolysis, heat
and photolysis. Such processes may toxify or detoxify a hazardous
substance or affect its ability to be transported.
  Persistence is also affected by the rate of transport to other en-
vironmental media: surface water, groundwater and air. Identifica-
tion and  quantification of  all surface water  courses,  ephemeral
and perennial, which cross  or drain the site  should be done  in-
itially. Surface drainage patterns and flow rates affect the erosion
of hazardous substances away from a site.  Flow rates, fluctuation
and pH may be measured in the field; in some cases, the data may
already be published water resource documents of State and Fed-
eral agencies or may be available through computerized data bases
maintained by the USGS and the USEPA.
  Needed data on groundwater include  the  identification of the
water table relative to the location of the wastes, connection to sur-
face water, descriptions of water bearing zones or aquifers located
at greater depths and quantification of these groundwater flow sys-
tems in terms of their flow rate, direction,  use, point of discharge
and pH. This information should be supplemented by a review of
the pertinent published literature  and maps on groundwater  re-
sources of the area surrounding the site. Sources of such informa-
tion include State  and Federal geological surveys  or equivalent
components within the state natural resource or environmental pro-
tection departments.
   Definition and evaluation of the groundwater flow  system may
require installation of sufficient monitoring wells on and about the
site to measure water table  elevations or potentiometric heads as
well as to provide access for groundwater quality  sampling.  In-
formation from these wells may be supplemented by existing mon-
itoring or water supply wells which may be located near the site.
Pumping records, particularly for large industrial or municipal pro-
duction wells, will yield important information on historic stress to
the aquifer and possible artificial  alteration of the  natural water
table or potentiometric surfaces.
  Another important component of the hydrologic system that can
affect  contaminant migration is precipitation and associated infil-
tration. An understanding of the seasonal range of precipitation
likely to fall at a site over an annual cycle is  essential to an accurate
determination of contaminant migration. Rainfall amounts can be
obtained through continuous recording rain gauges installed at the
site or from historic data collected by the National Weather Service
at the  nearest climatological station. Rainfall data can be used in
various USEPA and other models  to compute the amount of sur-
face runoff to be expected at a site and the  amount of water likely
to infiltrate the site and carry contaminants to the groundwater.
  The geologic environment, for the purpose of an endangerment
assessment, is considered to be the  soil and  subsurface materials in
which  or  upon which the hazardous waste may be located. This
medium is important in terms of its controlling effects on water
and contaminant transport. The physical and chemical characteris-
tics of the geologic setting should be described  and documented
using the results of on-site investigations supplemented by geologic
accounts in the published literature.
  Soil cover at the site should be defined in terms of depths, lateral
distribution and type.  This information  may be obtained by an
                                                       appropriate grid of sample cores collected on-site, together with in-
                                                       formation from county soil surveys which may be published for the
                                                       area by the USDA's Soil Conservation Service. Analyses of soil
                                                       samples will  yield physical  characteristics such as permeability,
                                                       grain size distribution and thickness as well as chemical characteris-
                                                       tics including contaminant concentrations and natural organic and
                                                       inorganic content. This information will be useful in  determining
                                                       presence and extent of soil contamination, availability of soil con-
                                                       taminants to infiltrating water, ability  of the natural soils to retard
                                                       contaminant migration and suitability  of existing soil as cover ma-
                                                       terials for remedial actions.
                                                         Descriptions of the subsurface unconsolidated  or consolidated
                                                       materials, including factors such as thicknesses, lithologies, water
                                                       bearing zones, hydraulic characteristics, strike, dip,  areal  extent
                                                       and continuity and engineering properties, may be assembled from
                                                       field investigation efforts and review of published  literature. Stud-
                                                       ies of regional geomorphology, bedrock and surficial geology and
                                                       mineral resource assessments  conducted by  universities, private
                                                       consulting firms  or State and Federal agencies all can provide val-
                                                       uable supplemental documentation of the critical geologic char-
                                                       acteristics  that may have a bearing on the ultimate transport  and
                                                       fate of contaminants migrating from a site.
                                                         Contaminants may also be transported through  the air by evap-
                                                       oration or suspension and subsequent deposition away  from the
                                                       waste site. Wastes can  be suspended or  adsorbed  on particulates.
                                                       The  distance particles will travel will  depend on  such factors as
                                                       particle size, precipitation, air turbulence, wind speed and topog-
                                                       raphy. Particle size also can determine the depth in the respiratory
                                                       tract that a substance is deposited. The final site of deposition of a
                                                       particle in the lungs can influence the  toxic effect  that may  be ex-
                                                       erted.
                                                         By combining knowledge of the likely movement, persistence
                                                       and transformation of a hazardous substance, it is possible to pre-
                                                       dict its partitioning and distribution among environmental media at
                                                       and away from a site.'  An evaluation of the fate and transport of
                                                       hazardous pollutants is crucial for developing an accurate estimate
                                                       of the levels and extent to which substances have been, or continue
                                                       to be, released into the environment. This estimate establishes the
                                                       foundation for an exposure assessment.

                                                       Estimates of Exposure

                                                         The identification and quantification of chemicals found or ex-
                                                       pected on and off the site can be related to the surrounding popula-
                                                       tions to estimate  exposures to the hazardous substances. Impor-
                                                       tant  factors to consider are population size and density, distance
                                                       separating the site from populated areas, accessibility to the site,
                                                       land use and recreational activity.
                                                         Exposure can be evaluated for the general population or for spe-
                                                       cific groups within the population. Certain high  risk subpopula-
                                                       tions should  be  identified,  such as  pregnant women,  children,
                                                       elderly or ill persons. Demographic information may be obtained
                                                       for specific Census Bureau tracts.
                                                         The toxic effects of some chemicals may be strongly influenced
                                                       by the route of exposure. Possible exposure routes include inhala-
                                                       tion of gases or dusts, absorption through the skin from direct con-
                                                       tact and ingestion of contaminated water, soil  or food. Popula-
                                                       tion characteristics and habits, as well as the amounts and concen-
                                                       trations of contaminants in various media, may affect the degree of
                                                       exposure by each route. The behavior  of the compound in a given
                                                       medium should also be considered, since some substances may be
                                                       bound and unavailable for absorption by the exposed persons.
                                                         The time pattern of exposure should be considered as  well. The
                                                       frequency,  duration and timing  of exposures may influence the
                                                       toxic effect that a hazardous substance  may exert.
                                                         Exposure to fish, wildlife  and domestic  plants and animals
                                                       should also be assessed. One important factor to consider is  the
                                                       bioconcentration of certain contaminants in the food web, which
                                                       can lead to increased exposures in the environment and to humans.
                                                       Another important consideration is the introduction of hazardous
214
FATE OF CONTAMINANTS

-------
substances into sensitive environments such as wetlands and habi-
tats for endangered plant and animal species.
Toxicity Evaluation

  In addition to exposure estimates, information on the toxicity of
the hazardous  substances at a site is needed to  perform an en-
dangerment assessment. Toxic effects may be expressed in gross
ways such as functional impairment, pathologic lesions or death, or
in less obvious  ways that may alter metabolism, immune response
or behavior.
  Physical and chemical properties of a substance, environmental
media in which it is carried, route of exposure, time pattern of ex-
posure and  personal characteristics of exposed individuals all may
influence the degree of toxic effect.
  For a toxic response to occur in a person, a sufficient quantity of
a substance must reach a receptor or target organ for a sufficient
period of time. The response depends on the dose of the chemical,
the length of exposure and transport to the receptor. Factors that
affect the absorption of the chemical and its distribution, metab-
olism and excretion, can profoundly alter its toxicity.
  Exposure to a toxic substance may produce immediate or delayed
effects. The onset and severity of effects may vary from organ to
organ and from person to person.  Toxic effects may be exerted at
the point of absorption, such as the skin, gastrointestinal tract or
lungs, or may occur in distant  organs  following  distribution,
metabolism or accumulation.
  A further complicating factor is that persons may be exposed to
several hazardous agents simultaneously or sequentially rather than
to just one. Such multi-chemical exposures may  result in altered
absorption, distribution and expression of effect.  Chemicals often
interact to produce effects either greater or less than the sum of
their individual effects. When such effects are known or suspected,
they should be considered in the toxicity evaluation.
  Resources are available to obtain established information on the
toxicity of certain substances. Controlled animal experiments using
acute, subchronic or chronic exposures may have been performed
to show toxic effects such as cellular or tissue changes, behavior
modification, carcinogenesis, mutagenesis, teratogenesis or death.
In  vitro studies using cell cultures or bacterial assays may have
demonstrated mutagenic or growth  process effects.  Such studies
provide insight into the possible toxic effects to  humans and the
mechanisms of toxicity. Evidence of toxic effects in humans from
exposure to hazardous substances may also be obtained from clin-
ical reports, anecdotal reports  or epidemiologic studies examining
associations between specific exposures and specific disease out-
comes.
  There are many difficulties in interpreting existing toxicological
data.  Animal and in vitro studies require caution when extrapo-
lating to  presumed effects in humans due to such differences as
body or organ size, system complexity, absorption and distribution
patterns or metabolic processes. However, these studies may be the
best, and often only, source of toxicity information available.
  Clinical reports often describe  acute or  chronic  effects from
large doses  of hazardous substances.  Unfortunately, such informa-
tion may not be representative of exposures at hazardous waste
sites, since those exposures may be long term and low level. Clinical
evidence can be suggestive of the kinds of effects that may occur at
hazardous waste sites but should be interpreted with caution.
  Similarly, health  related anecdotal  complaints of populations
near waste sites should be recorded and considered but should also
be  interpreted  with caution. Such information may be obtained
from  municipal or State public health departments or from the
U.S. Center for Disease Control.
  Epidemiologic  studies may or may not show an association be-
tween exposure to a chemical or mixture of chemicals and a par-
ticular disease  outcome. However,  results should be interpreted
with careful attention to study design and execution. Studies may
be  weakened by  selection  of improper comparison  groups, mis-
classification of disease or exposure status, recall bias, confound-
ing exposures to other hazardous substances or poor power to de-
tect an association. If these potential problems are minimized, valid
epidemiologic data can be the most useful evidence in assessing the
adverse effects of a toxic substance to humans.
  In addition to human health concerns, other adverse environ-
mental effects should be considered. Direct evidence of harm, such
as fish kills, recorded  declines in plant and animal populations or
decreased crop yields, may be related to a hazardous substance.
Adverse effects on the taste of drinking water or edible fish and
shellfish, the odor and clarity of air and other aesthetic problems
should also be considered.
ENDANGERMENT ASSESSMENT
  Evidence of endangerment at a specific site might include docu-
mented harm to the environment, increased reports of human ill-
ness or other observable adverse effects. Endangerment also may
be manifested in less evident ways or as threats of potential future
harm.
  To assess the endangerment posed by hazardous substances at a
waste site, expected exposures to surrounding populations and the
environment  are compared with toxicological information. The
merit of particular toxicological studies  and exposure  estimates
should be considered since the quality of data should determine
their relative importance in the endangerment assessment.
  However, scientists with similar training may have widely diver-
gent views on the interpretation of the data. These divergent views
include extrapolations from high dose to low dose, extrapolations
of animal data to humans, utilization of in vitro studies, the use of
data derived from  studies  involving one route of exposure to an-
other route and the health significance of certain biological effects.
  Studies of toxicologic effects and dose-response relationships can
be used for comparison to the estimated exposures at a specific site.
It is important to determine the circumstances under which these
numbers are applicable and the toxic effect for which they are de-
rived, since misapplication can lead to erroneous assessment of en-
dangerment. In addition, the data should be critically  reviewed for
methodological quality.
  Toxicological indices are generally species-specific  and refer to
concentrations or doses that result in specific toxic effects. These
indices may  be found in a variety of toxicological references.
Guidelines and standards for some hazardous substances have been
developed by government agencies and professional organizations.
These  numbers generally set a limit for the level of  a given sub-
stance in a particular medium that signifies no probable risk or a
certain degree of risk to human health or the environment.
  An endangerment assessment may be qualitative,  defining the
nature of potential or actual  harm  at a site. An assessment may
also be quantitative, attempting to define the degree of risk posed
by hazardous substances in terms of number of people affected and
damage to human health and the environment.
  For  non-carcinogenic substances, lifetime exposure  estimates
may be compared to "acceptable daily intake" (ADI) levels or
other standards appropriate to the situation." ADIs are assumed to
represent the dose  that is without probable risk when taken daily
for life. They are generally derived from doses which were observed
to have no toxic effect in animals, corrected by a safety factor that
reflects whether the data are from humans or animals,  the quan-
tity of data  and the existence  of  sensitive subpopulations. En-
dangerment is  assessed from a  comparison of the estimated ex-
posure and the ADI (or other guideline)  for a given  non-carcino-
genic substance.
  For  carcinogens, the  USEPA does not define a threshold re-
sponse level. Rather, any amount of exposure is expected to pose a
certain degree  of  risk. These risks are quantitatively  calculated
from the estimated magnitude of exposure and the potency of the
carcinogen, if known.
  Some aspects of the methodology for defining carcinogenic risks
are controversial, such as  the choice of model for risk estimation.
A number of models exist, using a variety of mathematical func-
                                                                                         FATE OF CONTAMINANTS
                                                                                                                             215

-------
tions, each with a biological basis used to justify its form. The main
difference among these models  is the shape of the dose-response
relationship as the exposure decreases.
  Until November  1980,  the  USEPA Cancer  Assessment Group
(CAG) used a linear, non-threshold, one-hit model to estimate can-
cer risk.5 Following public comment, CAG later adopted a linear-
ized, non-threshold, multi-stage model for the  extrapolation since
it  appears  to  be the  most  general  and  biologically  plausible
method.'  The calculated  carcinogenic risk  is the  probable  upper
bound of risk. The estimate is conservative,  and the true risk is not
likely to exceed it.
  The  qualitative  and  quantitative endangerment  assessments
developed through  this process  are valuable in defining the kind
and degree of risk posed  by hazardous substances at a waste site.
They are also useful in providing guidance for remedial alternative
selection and for justifying that choice.
  This discussion of endangerment assessment has been necessarily
general. The wide variety  of chemicals found from site to site and
the range of site conditions  require site-specific assessments that
preclude more detailed guidance.
  The adequacy  of, and  confidence in, an endangerment assess-
ment is highly dependent on the  quality of sampling, analyses, fate
predictions, exposure estimations and toxicology data. At all stages
of an assessment, careful  attention must be paid to the methodol-
ogy and executive of information collection. Interpretation of this
information involves a great deal of judgment and should reflect
sound scientific principles and a commitment to the protection of
public health and the environment.
CONCLUSIONS
  An endangerment assessment at a hazardous waste site should
integrate knowledge of the toxicological effects of hazardous sub-
stances with the estimated environmental and human exposures to
these substances. Exposures are  estimated based on site character-
                                                         istics, contaminant properties, hydrogeological conditions and en-
                                                         vironmental measurements. Strict quality control is  important in
                                                         the collection of data, and sound scientific judgment  is required
                                                         in interpretation. The USEPA performs these assessments to de-
                                                         termine appropriate emergency or remedial responses and to justify
                                                         these actions.
                                                          REFERENCES

                                                          1. Anderson, E.L., "Quantitative methods in use in the United States to
                                                            assess cancer risk," presented at the Workshop  on Quantitative Esti-
                                                            mation of Risk to Human Health from Chemicals, Rome, Italy, 1982.
                                                          2. Falco, J.W., Mulkey, L.A. and Schaum, J., "Multimedia modeling
                                                            of transport  and  transformation of contaminants," in  Environment
                                                            and Solid Wastes.  Francis, C.W. and Auerbach, S.E., Eds., Ann Arbor
                                                            Science, Ann Arbor, MI, 1983.
                                                         3. Federal Register 44. (52) (144)  (191). Mar.  15,  July 25, October  1,
                                                            1979,  "Environmental Protection Agency Water  Quality  Criteria,"
                                                            15926-15981, 43660-43697, 56628-56656.
                                                          4. Federal Register 45,  (231), Nov. 28, 1980,  "United States Environ-
                                                            mental Protection Agency," Water Quality Criteria Documents; Avail-
                                                            ability," 79350-79353, 79379.
                                                          5. Federal Register 47,  (137), July 16,  1982, "National Oil and Haz-
                                                            ardous Substance Contingency Plan, Rules and Regulations," Subpart
                                                            F, 31213-31218.
                                                          6. Lyman,  W.J.,  Reehl, W.F. and Rosenblatt, D.H.,  Eds. Handbook
                                                            Chemical Properly Estimation Methods. McGraw-Hill, New York, NY,
                                                            1982.
                                                          7. Neely, W.B.,  Organizing data  for environmental studies.  Environ.
                                                            Toxicol. andChem. I, 1982, 259-266.
                                                          8. Public Law 96-510,  "Comprehensive Environmental Response, Com-
                                                            pensation,  and Liability Act of 1980," 94 STAT. 2767-2811, Dec. 11,
                                                            1980.
216
FATE OF CONTAMINANTS

-------
             MIGRATION  AND DEGRADATION  PATTERNS
                    OF  VOLATILE ORGANIC  COMPOUNDS

                                             PATRICIA V. CLINE
                                               DANIEL R. VISTE
                                            Warzyn Engineering Inc.
                                               Madison, Wisconsin
 INTRODUCTION

   Volatile  organic  priority pollutants have  been detected  in
 groundwater all across the country. These compounds, widely used
 as solvents, are considered environmentally mobile and persistent.
 Improved analytical methods using  gas chromotography and/or
 mass spectroscopy now allow their detection at extremely low
 levels. The presence of the synthetic organics in groundwater coupl-
 ed with an improved ability to detect them has resulted in increas-
 ing numbers of contamination investigations.
   Biodegradation is not typically an integral part of today's
 groundwater  investigations. There  is considerable controversy
 regarding whether degradation is an important factor in determin-
 ing the  fate of the chlorinated volatile organic priority pollutant.
 Increasing evidence indicates chlorinated solvents can be degraded
 in an anaerobic environment by reductive dehalogenation.  It is
 reported this process occurs when the oxidation/reduction poten-
 tial is less than 0.35V. The sequential removal of chlorine atoms
 from halogenated 1  and 2 carbon aliphatic compounds results in
 formation of other volatile, chlorinated priority pollutants which
 can be detected during investigations of solvent contamination.li2'3
   In this paper, the authors present data from a variety of sites hav-
 ing documented chlorinated solvents contamination. Three types of
 sites were selected to illustrate breakdown patterns which  may
 develop as a result of diverse environmental conditions.  Data from
 landfills are presented to demonstrate presence of degradation pro-
 ducts in  biologically active anaerobic environments. Two solvent
 recovery facilities  which handle  both  chlorinated  and  non-
 chlorinated solvents showed similar migration and degradation pat-
 terns. Finally, an industrial site with no apparent degradation
 demonstrates  conditions in which reductive dehalogenation  may
 not be a primary degradation mechanism.
  Research data indicates chlorinated solvents have varying rates
 of breakdown. The data were therefore evaluated for a dominance
 of compounds which  show longer half-lives, including  1,2-di-
 chloroethenes and vinyl  chloride.4
 BACKGROUND
  For purposes  of  this evaluation, selected compounds were
 designated as "parent" compounds based on their widespread use
 and/or  known presence at these  specific sites. They include
 methylene chloride, 1,1,1-trichloroethane, trichlorethene and tetra-
 chloroethene.
  Breakdown products are designated as compounds which would
 result from reductive dehalogenation of these parent compounds
 and include dichloroethanes, chloroethane, dichloroethenes  and
vinyl chloride. For purposes of this evaluation, methylene chloride
is disregarded, since it  is a commonly used solvent, potential
degradation product and common laboratory contaminant. Em-
phasis is placed on the ethene and ethane series because there is less
ambiguity in the assignment of parent and breakdown products.
The anaerobic breakdown sequence for the chlorinated ethenes and
ethanes via reductive halogenation is shown below:

Chlorinated Ethenes

              Ci             Cl          Cl
              T                cis-1,22
              I              *          T
Tetrachloroethenei—> Trichlorethenej!—>trans-l,22 —»V1nyl Chlorldej '^'

                               1.1-1
                              dichloroethene
Chlorinated Ethanes
1,1,l-Tr1chloroethane2
1,1-dichloroethanei
                                      Cl
                                      T
chloroethanei (2)
1. Research indicates substantial degradation.
2. Research indicates degradation is slow.

  In work performed at the Florida International University by
Parsons,  Wood  and  DeMarco,'  biodegradation  of  either
trichloroethene or tetrachloroethene produced higher concentra-
tions of cis- 1,2-dichlorethene when compared to the trans-isomer.
  Trans-1,2-Dichloroethene is a priority pollutant and has a lower
allowed concentration in drinking water (272 /*g/l) than the cis-
isomer (400 /tg/1).5 USEPA's rationale for selection of the trans-
isomer as the priority pollutant was based on the availability of the
analytical standard."
DATA PRESENTATION

  In the authors' first attempts to correlate the ethene breakdown
series with data from contaminated sites, it became apparent that
the dominant dichloroethene compound detected was trans-l,2-di-
chlorethene.  The cis-isomer  is  not  a  priority  pollutant and,
therefore, is not mentioned in  the methods  for  analysis of the
volatile organic priority pollutants using Method 601 or Method
624.
  These methods recommend the use of a column composed of 1%
SP 1000 on Carbopack B. The isomer pair cannot be separated us-
ing the above column. In addition, since they have identical mass
                                                                                  FATE OF CONTAMINANTS
                                                      217

-------
spectra, the isomer  pair will  not  be differentiated by mass spec-
trometry and will subsequently be identified as the trans-isomer.
  The above theory was verified by the submittal of a standard mix
containing  both  the  cis-  and  trans-isomers to  a  prominent
midwestern laboratory. Analysis  by Method 624 found only the
trans-isomer, but the quantitated result equalled the known total of
the isomer pair.
  The Michigan Department of Health can separate the cis- and
trans-isomers and, in a current  investigation, has determined that
the major contaminant at a site is not trans-1,2-dichloroethene as
found by a USEPA contract laboratory, but is the cis-isomer. They
have indicated  that frequently they find the cis-isomer and, if con-
centrations are high,  they  occasionally find traces  of the trans-
isomer.
  Based on this information, the authors conclude that much of
what is typically reported as  the  trans-isomer, which is a priority
pollutant, is  in fact cis-1,2-dichloroethene.  In  the  subsequent
evaluations, the authors  will refer to these compounds as 1,2-di-
chlorethenes.
Landfills
  Landfills which accept municipal waste  have an anaerobic en-
vironment in which substantial  breakdown of compounds occurs.
At  sites  which  have  also  accepted waste products containing
solvents, a number of volatile  organic  priority pollutants can be
detected in the leachate. The analyses of five leach ate samples from
Site #1 which  accepted both municipal and industrial wastes are
found in Table  1. The site also received significant quantities of
hazardous and nonhazardous liquid wastes. Based on records of
waste accepted, there is a dominance of "breakdown products" at
this site.
  The amount of breakdown products detected in groundwater at
two other sites where volatile organic contaminants have migrated
off-site is shown in Table 2. Site #2 is a small municipal landfill in a
sand and gravel environment and Site #3 is a large clay-lined site
which has accepted waste  similar to Site  tfl. At these sites, the
authors have also documented a dominance of the breakdown pro-
ducts in groundwater downgradient from the waste disposal boun-
daries.
  The purpose of presenting leachate data from these landfills is to
demonstrate that in an anaerobic, high-organic matrix, one is likely

                             Table 1
                     Landfill  Leachate, Site #1

                                     Leachste Sample Number
   Chlorinated Ethanes

      1 Trlchlorethanes
           2  1,l-D1chloroethane
             l,<:-01chloroethane
             Chloroethane
   Chlorinated,Ethanes

      1  Tetrachloroethene
        Trlchloroethene

          2 l,2-D1chloroethenes
            1,l-D1chloroethene
            Vinyl  Chloride
                            ND3  68   ND  NO  NO

                         1.500  240  130  11  13
                            ND
                            NO
12
21
                            ND   13
                            NU  100
21   ND  ND
18  160  ND
     NO  ND  ND
     62  NO  ND
                         3,200  990  950 150  ND
                           ND   NO   ND  ND  NO
                           NU  120   59 100  NO
   Otner_Volat1le  Priority Pollutants

       Methylene  Chloride           5,300  120   770  ND  14
       Toluene                     2,000  410   600 460  58
       Benzene                        ND   30    37 110  16
       Ethylbenzene                    HD   93    64 140  68
       1,2-Dichloropropane             ND   18    37  ND  ND

  I. Parent Compounds   3. ND — {10 ^g/l
  2. Breakdown Products 4. All Concentrations In Mg/l
                                                                                      Table 2
                                                               Breakdown Products Present In Contaminated Groundwater
                                                                               Near Disposal Facilities

No. of Samples from Wells showing
Solvent Contamination
No. of Samples with 50% Breakdown
Products*
No. of Samples with 50-75% Breakdown
Products*
No. of Samples with 75- 100% Breakdown
Products*
Sltett

10

2

3

5
site in

8

0

0

8
                                                          •Breakdown defined as monochJoro- and dichloro- ethane* and elhenes compared u> total
                                                          chlorinated ethanes and ethcnes

                                                          (o find compounds which are a result of reductive dehalogenation.
                                                          It  is unlikely that these compounds were the dominant disposal
                                                          compounds at these sites based on site records, general production
                                                          and common use. Of particular interest is the fact that all eight of
                                                          the leachate samples from the large co-disposal facility were com-
                                                          prised of greater than 75% breakdown products.

                                                          Solvent Recovery Facilities

                                                            Solvent recovery facilities handle a wide variety of organic com-
                                                          pounds  including  chlorinated solvents.  In addition, varying
                                                          hydrogeologic conditions can result in complex migration patterns.
                                                          The two facilities discussed in this section differ in operation and
                                                          location, but have  similarities  in migration and degradation pat-
                                                          terns. Geologic and  hydrologic  characteristics  at  these solvent
                                                          recovery facilities are given in Table 3.

                                                                                      Table 3
                                                                       Solvent Recovery Site Geological Conditions
                                                          Location
                                                          Dale of
                                                          Investigation
                                                          Geology
                                                          Hydrology
                                      Site 1

                                      Connecticut

                                      1980
                                      Alluvial sands and gravel in
                                      relatively impermeable bedrock
                                      valley
                                      Shallow groundwater.  10 ft.
                                      alluvial sands constitute primary
                                      municipal aquifer
                                                                                                 S1U2

                                                                                                 Wi
                                                             198}
                                                             Thick, sandy glacial till deposits
                                                             overrying limestone bedrock


                                                             Till supports only minimal ground-
                                                             withdrawal, permeability appro*
                                                             IO-4 to 10-' on/sec. Limestone
                                                             is aquifer in the area.
  The analytical data for the above sites are found in Tables 4 and
5.  Both  sites handled  chlorinated and nonchlorinated  solvents.
High  concentrations of both  the chlorinated and nonchlorinated
compounds were present near the on-site handling areas. The off-
site contamination  showed a  dominance of the chlorinated com-
pounds.   Nonchlorinated  compounds   detected  were   priority
pollutants. In cases  where analyses were performed, the presence of
compounds like toluene and  benzene were  indicative of a much
higher concentration of other  nonpriority pollutant hydrocarbons.
  At  the Wisconsin site, dichlorethanes, dichlorethenes and vinyl
chloride were detected in significant concentrations in the ground-
water. These compounds were not handled at the  facility, and this
is supported by records of routine gas chromatographic analyses at
the recycling facility. Further  evaluation  failed  to  indicate  the
presence of other possible sources of the breakdown  products. In-
formation  was not  available to  evaluate this question at  the Con-
necticut site.
  An evaluation was  then performed to assess whether data from
these  facilities show patterns which would be a result of anaerobic
degradation. The evaluation includes an analysis of the percentage
of breakdown products measured  at  the source  and at a down-
gradient  location.
  To  illustrate trends, the data have been summarized in  Figure I.
Results are shown  for the priority pollutant analyses for a water
 218
FATE OF CONTAMINANTS

-------
table well and piezometer located on-site that had the highest con-
centrations, as  well  as a  downgradient  water table well and
piezometer. At both  of the sites, primarily horizontal hydraulic
gradients were observed during the hydrogeological assessment
based on water level measurements. Elevated concentrations  of
contaminants  were anticipated  at the downgradient water table
wells.
                            Table 4
                   Solvent Recovery Operations
          Summary of Volatile Organic Priority Pollutants Detected at
           On-site and Downgradient Piezometers: Connecticut Site

On-S1te
Water Table At
Chlorinated Ethanes
1 1,1,1-Trichloroethane
2 1,1-Oichloroethane 8
Chlorinated Ethenes
1 Tetrachloroethene 2
Trichloroethene 39
2 l,2-D1chloroethenes 30
1,1-Dichloroethene
Vinyl chloride
Other Solvents Detected
Methylene Chloride 100
Ethyl benzene 12
Toluene 34
N03 3,
,300 - 3,
,900
,000
,000 2,
ND
NO
.000 7,
.000
,000 5,
250'
Downgradient
Depth
700
000
ND
330
700
NO
200
000
440
100
Water
Table
260
2,500
34
NO
ND
ND
NO
25
ND
ND
At
Depth
ND
NO
NU
ND
4,300
NO
2,700
3,900
3,700
7,600
1. Parent Compounds 3. ND — <10/ig/l
2. Breakdown Products 4. All Concentrations are in j»g/l
                           Table 5
                  Solvent Recovery Operations
        Summary of Volatile Organic Priority Pollutants Detected at
          On-Site and Downgradient Piezometers: Wisconsin Site
On-S1te
Water
Chlorinated Ethanes
1 1,1,2,2-Tetrachloroethane 19,
1,1,2-Trichloroethane
1,1,1-Trichloroethane 22,
2 1,2-Dichloroethane
1,1-Dichloroethane
Chloroethane
Chlorinated Ethenes
1 Tetrachloroethene
Trichloroethene 63,
2 1,2-Dichloroethenes 30,
1,1-Dichloroethene
Vinyl Chloride
Other Solvents Detected
Methylene Chloride 230,
Benzene 12,
Ethyl benzene 28,
Toluene 100,
Table
QUO
ND
000
ND
ND
NU
ND
000
000
ND
ND
000
000
000
000
Depth
N03
ND
270,000
ND
6,200
ND
22,000
250,000
8,700
ND
ND
170,000
ND
9,200
42,000
250'
Downgradient
Water
Table
NU
ND
ND
ND
ND
ND
ND
NU
ND
ND
ND
ND
!!0
ND
ND
Depth
ND
60
20,000
230
5,100
90
610
1,000
47,000
720
210
20,000
20
630
4,100
1. Parent Compounds 3. ND — (lOw/l
2. Breakdown Products 4. All Concentrations are in jig/1
  The figure shows the  total  volatile  organic concentrations
detected at the above described well locations for both sides and the
percentage of breakdown products compared  to the sum of the
chlorinated ethanes and ethenes.
  Both of the sites exhibited high levels of chlorinated organic con-
tamination at the source. Nonchlorinated organics were also pre-
sent at the  sources  in  high concentrations,  providing a  non-
chlorinated carbon source. These nonchlorinated  organic com-
pounds were present in highest concentrations  at the water table.
At the Wisconsin site, a floating layer of fuel oil type material was
detected at one well.
  With distance downgradient from the source, the contaminants
were detected at greater concentrations with depth  even though
groundwater flow was near horizontal. There are various explana-
tions for this phenomenon,  including changing groundwater  flow
SOL
iimnmBnvi
48t :
rce


200 ppm



591 :
Do
milll E Illlls Hill Slllllfe
v
89%:
unqr


3 ppm
FLOW
DIRECTION

100J:
adient
mmm

Monitoring Well
(typical)


Well Screen
(typical)
25 ppm 25 ppm
CONNECTICUT SITE
So
lllllslll||iŁ)|||l
jrce

22Xi
500 ppm
800 pj
Do
Hill ~!1I|!^»|II»'SJ (HI'S.
ND :
FLOW
DIRECTION
mgr


m 100 [
WISCONSIN SITE
idlent
H|||m)IUI»

pm
                                                                   In Values = % Breakdown Products  ppm Values = Total of all Volatile Priority Pollutants

                                                                                              Figure 1
                                                                            Anaerobic Breakdown Patterns of Organic Materials

                                                                   patterns,  recharge or  impermeable barriers  which  may have
                                                                   hampered migration of  contaminants to the water table  wells.'
                                                                   These  parameters will  be  evaluated  further  with  additional
                                                                   hydrogeologic study, where funding is available.
                                                                     Other explanations include  density effects, volatilization and
                                                                   selective degradation. It is well documented that chlorinated com-
                                                                   pounds will sink in the aquifer at the source when in excess of the
                                                                   solubility of water.7 For  subsequent density effects to be apparent
                                                                   in the contaminated groundwater where concentrations are lower,
                                                                   the overall density of that solution must be greater than that of
                                                                   background water quality. Preliminary calculations indicate that at
                                                                   the concentrations measured  at the sites, the density difference
                                                                   would not be sufficient to account for sinking of the contaminated
                                                                   groundwater plumes.
                                                                     The USEPA had indicated that a primary environmental fate for
                                                                   these compounds in aquatic systems would be volatilization.1 Fac-
                                                                   tors which affect volatilization of these compounds from a ground-
                                                                                           FATE OF CONTAMINANTS
                                                           219

-------
water system include: soil porosity and temperature, depth to water
table and  the  various  solubilities  of the compounds in water.
Although it is recognized that some volatilization will occur, it is
not expected to be a primary fate of organics at these sites.
  Selective degradation is  presented as another possible explana-
tion for preferential loss of the constituents at the water table wells.
The biodegradation of chlorinated compounds may be affected by
the  co-metabolism  of  other carbon sources. Solvent recovery
operations can provide a nonchlorinated carbon source which tends
to accumulate near the water table surface. These compounds are
typically not detected with distance from the source, due to rapid
breakdown and  may be responsible for preferential loss of the
chlorinated compounds from the more shallow  zone of the aquifer.
  The breakdown of the chlorinated compounds can occur rapidly
in the presence of a nonchlorinated carbon source which promotes
rapid co-metabolism to dehalogenate the chlorinated compounds.
The data suggest that degradation continues to occur deeper in the
aquifer, perhaps at a slower rate.

Industrial  Site
  For purposes of contrast with sites which have high levels of con-
tamination and  a substantial  carbon  source, the  authors  have
presented data from an industrial site having primarily sandy soils,
shallow groundwater and little or no  detectable nonchlorinated
organic priority pollutants (Table 6).

                            Table 6
         Industrial Site Solvent Contamination of a City Well
Wtl] l.l.l-Trichlorocthint
1 ND
2 13,800
3 2,660
4 7
5 8
6 ND
7 10
Trkhlorovthene
81
2,040
410
1
2
68
12
l.l-Dichloroctliene
ND
250
ND
ND
ND
ND
ND
1 . All Concentrations are in pg/1
                           2. ND — ( I fig/I
  Three major contrasts  with data from the solvent recovery
facilities are noted:
•Overall contaminant concentrations detected are lower and  all
 compounds are chlorinated
•A dominance of the parent compounds exists
•The plume was detected  in highest concentrations at the water
 table wells. The lack of a significant carbon source to promote
 degradation can account  for  the minimal breakdown occurring
 at the industrial site

CONCLUSIONS
   Parameters which would assist in determining biodegradation ac-
tivity are typically not incorporated into standard hydrogeologic in-
vestigations. A better understanding of  the  role of degradation
could be obtained through a more comprehensive investigative pro-
gram  including biological  assessment  as  well as  the standard
groundwater flow and chemistry analyses
   Data from the authors' investigations suggest that if a site has a
substantial  carbon  source,  anaerobic  degradation will occur,
resulting in the formation of dichloro-  and monochloro- ethane
and ethene compounds. The presence of these compounds follows
the predictions in the literature regarding the degradability of  the
parent compounds. In addition, the dominance of the cis-isomer of
1,2-dichloroethene  formed during degradation will result in its
presence in these investigations rather than the priority pollutant
trans-isomer.
   A floating organic layer  near a contamination site may enhance
the rate of degradation near the water table as the chlorinated com-
pounds would more readily be co-metabolized in that zone of the
aquifer.

RECOMMENDATIONS

  At sites where degradation is indicated, additional measurements
should be made to better understand the potential role and con-
trolling  mechanisms  of biodegradation.  This  would  include
measurement of the overall organic content in water or soil and
measurements of oxidation  reduction potential (Eh), oxygen con-
centration  and  plate counts  of  bacteria."•I2>"'14'15  Density
measurements of  the contaminated  groundwater will allow
clarification of  potential density effects on  migration patterns.
During interpretation of the data, one can evaluate the presence of
breakdown products and the pattern of their occurrence in relation
to the  parent compounds.  One  should report  " 1,2-dichloro-
ethenes" without specifying the specific cis- or trans-isomer, unless
that specific distinction can  be made by the analytical laboratory.
  It  is hoped that  increased  awareness of the conditions under
which maximum degradation can occur will improve the approach
and substantially increase the conclusions which can be drawn from
groundwater contamination investigations.

REFERENCES

 I. Parsons,  F., Wood, P.R. and DeMarco, J., "Transformations of
   Tetrachloroethene and Trichloroethene  in Microcosms and Ground-
   water," J. AWWA, Feb.,  1984, 56-59.
 2. Wood, P.R.,  Lang, R.F.  and Payan, I.L.,  Anaerobic Transforma-
   tion,  Transport and Removal of Volatile Chlorinated Organics in
   Groundwaler. Drinking Water Research Center. School of Technol-
   ogy,  Florida  International University, Tamiami Campus, Miami,
   FL, 1981.
 3. Kobayashi,  H . and Rittmann, B.E., "Microbial Removal of Haz-
   ardous Organic Compounds."  Environ.  Sci.  Technol.,  16,  1982,
   170A-182A.
 4. Tabak, H.H., Quave, S.A.,  Mashni, C.I. and Barth, E.F., "Biode-
   gradability  Studies with Organic Priority Pollutant Compounds."
   JWPCF,  53, 1981, 1503-1518.
 5. Weiss, H.,  Status of DHSS Recommendations for Drinking Water
   Organic Chemical Contaminants Interim Health Advisory Opinions,
   1984.
 6. Bruehl, D.H., Chung, N.K.  and  Diesl, W.F., "Geologic Studies of
   Industrially-Related Contamination: Soil and  Groundwater Investi-
   gations." Proc. National Conference on Management of Uncon-
   trolled Hazardous Waste Sites, Washington. DC. Oct., 1980.
 7. GeoTrans, Inc., RCRA Draft Permit Writers Manual: Groundwater
   Protection,  1983. 3.14-3.15.
 8. USEPA.  Water Related Fate of 129 Priority Pollutants, Vol. 1, 38.1-
   53.13.
 9. Mackay,  D., Bobra, A., Chan, D.W. and Shiu, W.Y.,  1982. "Vapor
   Pressure Correlations for  Low-Volatility Environmental Chemicals."
   Environ.  Sci. Technol., 16. 1982, 645-654.
10. Wilson. J.T.. Enfield, C.G., Dunlap,  W.J.,  Cosby,  R.L.. Foster,
   D.A.  and Baskin,  L.B.,  "Transport and Fate of Selected Organic
   Pollutants in a Sandy Soil," J. Environ. Quality, 10, 1981,  501-506.
11. Wilson, J.T.. McNabb, J.F., Balkwill, D.L.  and  Ghiorse, W.C.,
   "Enumeration and Characterization of Bacteria Indigenous to a Shal-
   low Water-Table Aquifer," Ground Water. 21,  1982, 134-142.
12. McNabb, J.F.  and Dunlap. W.J.,  "Subsurface Biological Activity
   in  Relation to Groundwater Pollution,"  Ground Water,  13,  1975,
   33-44.
13. Ehrlich. G.G., Goerlitz, D.F., Godsy, E.M.  and Hull, M.F.. "De-
   gradation of Phenolic Contaminants in Groundwater  by Anaerobic
   Bacteria: St. Louis Park, Minnesota," Ground Water,  20, 1982, 703-
   710.
14. Allen, M.J.  and  Geldreich, E.E., "Bacteriological  Criteria for
   Groundwater Quality," Ground  Water,  13, 1975, 45-52.
15. Dul, E.F., Fellman, R.T. and Kuo, M.F., Evaluation of Systems to
   Accelerate Stabilization of Waste Piles or Deposits. Envirosphere
   Company, Lyndhurst, NJ. Roetzer, J.F.  Woodward-Clyde Consul-
   tants, Wayne, NJ, 1984, 109-204.
16. USEPA Effluent  Guidelines Section, personal communication.
220
        FATE OF CONTAMINANTS

-------
                       SITE ASSESSMENT UNDER CERCLA:
         'THE IMPORTANCE OF DISTINGUISHING HAZARD
                                              FROM RISK'

                                   SWIATOSLAV W. KACZMAR, Ph.D.
                                        EDWIN C. TIFFT, JR., Ph.D.
                                   CORNELIUS B. MURPHY JR., Ph.D.
                                        O'Brien & Gere Engineers, Inc.
                                              Syracuse, New York
INTRODUCTION

  The  Comprehensive  Environmental Response,  Compensation
and Liability Act of 1980 (CERCLA) otherwise  referred to as
"Superfund"  was established to provide a mechanism  for the
timely identification and remediation of hazardous substance re-
leases. The Act functions by providing a fund for  the cleanup of
waste emergencies and makes provisions for the allocation and re-
covery of resources regardless of the existence or willingness of par-
ties responsible for the release. CERCLA places government agen-
cies or their assigned designees  in a lead  position investigating
hazardous waste  sites  and emergencies to specify remedial re-
sponses and initiate their implementation.
  CERCLA and its procedures document, the National  Contin-
gency Plan, specify that response to implemented  in cases where
an imminent and substantial threat to health, human welfare and
the environment has been determined to exist.  Remediation is to
be conducted in a cost-effective manner focused on reducing risks
to acceptable levels. The working and intent of Superfund indicate
that  responses are to be directed toward actual  site risks,  and
acknowledges the existence of acceptable levels of risk.
  Crucial to the effective implementation of CERCLA is  the dis-
tinction between chemical hazards and site risks. Whereas chem-
ical hazard is an inherent characteristic such as toxicity, or reactiv-
ity or environmental persistence, risk  is a site or incident-specific
probability term. Site risk considers chemical hazards and a num-
ber of site related hazards, and represents the probability of a given
deleterious effect occurring. Many improperly  disposed chemical
compounds can be extremely toxic but become  so strongly bound
to soil particles or too rapidly degraded that they constitute no sub-
stantial risk to nearby populations. In other cases, a toxic chem-
ical may be released from a site but represents a very low risk due
to the absence of exposed human or wildlife populations in zones
where toxic levels of the  material exist. Both cases represent uncon-
trolled hazardous chemical waste conditions with a low probability
or risk of causing harm.
  A  specific charge of the National  Contingency Plan was the
compilation and publishing of a graded list of uncontrolled chem-
ical waste sites requiring immediate  attention  under CERCLA,
known  as the National  Priorities List. Most sites on the list were
compiled using a hazard ranking system known as the MITRE
model detailed in the National Contingency Plan.  It is important
to note  that the MITRE model is a generic method  of ranking site
hazards using a minimum of information and is not indicative of
actual risks of harm to human health or the environment. Inclusion
on the list indicates a situation that, due to some of its hazard char-
acteristics requires more detailed attention. Further assessment is
conducted during the site  Remedial  Investigation/Feasibility
Study.
THE REMEDIAL INVESTIGATION
  The determination of actual risks arising as a result of a given un-
controlled hazardous waste situation is the product of the Remedial
Investigation. The risk assessment may be the most useful criterion
used in determining the level of remediation, if any to be imple-
mented. Elements of a Remedial Investigation contributing to an
accurate risk estimate are a thorough characterization of the waste
materials deposited at the site, information on the physical and
hydrological characteristics of on-site and offsite areas and accu-
rate analytical chemical data  on the levels of contaminants of con-
cern in media leaving the site. The establishment of risk requires a
source of toxic chemical, a working mechanism which transports
the chemical  from the site to offsite areas and the existence of
potential for exposure of human  or wildlife  populations. The
source,  transport and affected population or "receptor" compon-
ents each represent a hazard term. No single hazard component
can stand alone to describe the level of risk arising as a consequence
of a waste site. The three components must act in concert before
the situation is such that it represents an imminent risk requiring
response under CERCLA.

INHERENT WEAKNESSES OF COMPREHENSIVE
ANALYTICAL SURVEYS
  In many investigations, activity begins by performing an analyti-
cal chemical  survey  of the  site and some offsite areas.  This
approach attempts to unify the source, transport and receptor com-
ponents of risk by demonstrating that chemicals are present in all
three compartments. Very often the objective of  the survey is to de-
tect and identify as many chemical components as possible. How-
ever, complete characterization of a given environmental sample
by non-specific analytical chemical techniques  is a difficult task
which cannot address all classes of compounds with an equal de-
gree of accuracy. Problems arise as a result of the large number of
chemical compounds and matrix interferents which may be present
in the waste mixture or environmental sample. Differences in their
relative  concentrations extractability,  recovery and  chromato-
graphic behavior during various cleanup, concentration, and sepa-
ration steps of the analysis make detection and accurate determina-
tion of each component unlikely.
  Even analytical protocols utilizing elaborate mass spectral li-
braries  can only detect and  identify a fraction of  chemical com-
pounds. The  EPA Priority Pollutant scheme is one of the most
sophisticated  multi-component analytical protocols in routine use
today, yet it can provide a false sense of security to investigators
                                                                            ENDANGERMENT ASSESSMENT
                                                     221

-------
              uu.o
               R1C
                                                                                                      1648
                                                                                    —I	
                                                                                     HBO
                                                                                     I7l30
                                                                                        	1	•	
                                                                                         1600       SCAH
                                                                                         20iOO       TIME
—I	
 400
 5:00
—\	
 GOO
 7i30
—I	
  000
 10.00
1000
I2i30
—I	
 1200
 ISiOO
                                                           Figure 1
                                             Gas Chromatogram of Base-Neutral Extract
 utilizing it as a criteria for site characterization. Figure 1 is a repro-
 duction of  a base neutral priority pollutant  scan of  a ground-
 water sample collected near an uncontrolled waste site. The sample
 contained over 1,000 parts per million of Total Organic Carbon
 (TOC). The analytical report indicated that no priority pollutants
 were present at a limit of detection in the low part per billion range.
 Yet it is evident that a significant number of chemical compounds
 have been detected. Computer matching of the mass  spectra of
 each peak yielded only two identifications with an acceptable prob-
 ability of positive identification. Further identification is limited
 because few analytical laboratories providing  mass spectroscopic
 services are  able to furnish manual interpretations of mass spectra
 within reasonable  constraints of technical validity, time and cost.
 Even if complete chemical characterization were possible, the in-
 vestigator would be left with a set  of chemical compounds whose
 hazards need to be evaluated.
 FORMULATION OF SITE-SPECIFIC PARAMETERS
   In light of the limitations of Priority Pollutant  scheme and the
 absence of an analytical technique  capable of detecting, quantify-
 ing and providing  comprehensive identification of components in
 a waste mixture or environmental sample, the remedial investigator
 might be well justified in reducing his dependence on the analytical
 chemist as a provider of primary data for use in preliminary  inves-
 tigations. An alternate approach is a thorough review of the site
 history and  activities with the intent of formulating a set of site-
 specific indicator compounds. This set is to represent  those ma-
 terials whose combination  of physicochemical and toxicological
 properties are unique from the other waste components in a way
 that establishes biological consequences of exposure and facilitates
 transport by the movement of air or water.
   The site-specific compound review  is  favored because it  ad-
 dresses  the source, transport and toxicological hazard character-
 istics of the  waste components and does not require a comprehen-
                                                        sive analytical search for the entire set of materials handled on the
                                                        site. It narrows the field of chemicals to a few compounds with a
                                                        known identity for which specific analytical protocols can be de-
                                                        signed and carried out within acceptable quality control criteria;
                                                        yet it does not limit itself to the Priority Pollutants.
                                                          The formulation  of a site-specific indicator  compound list is
                                                        initiated by careful review of the history of the site. This approach
                                                        is most successful when investigations are conducted with the co-
                                                        operation of site owners and site history information is available.
                                                        Review consists of collecting and interpreting information on pro-
                                                        duction processes, the raw materials and likely waste products of
                                                        the processes, their volumes and the waste disposal practices em-
                                                        ployed on the site. The product of the initial task is a list of chemi-
                                                        cal compounds most likely to have been disposed or otherwise dis-
                                                        tributed about the site through production activities.
                                                          The list of materials is then evaluated for toxicity and physico-
                                                        chemical characteristics which can, in combination with a transport
                                                        mechanism such as groundwater movement, function to facilitate
                                                        transport to receptors. Chemical characteristics  evaluated are the
                                                        materials reported acute  and chronic toxicity  and key physico-
                                                        chemical constants controlling environmental fate and mobility,
                                                        such as, vapor pressure, water solubility, octanol-water partition
                                                        coefficient  (!(„,), chemical reactivity and biodegradability. Data
                                                        on these parameters can be retrieved from chemical handbooks'-2
                                                        and primary references as  well as calculated by  physical chemical
                                                        property estimation techniques.'
                                                          The toxicity of a component is a term which initially stands alone
                                                        for evaluation.  Toxicity information can be retrieved  from stan-
                                                        dard reference books,' or by use of databases such as DIALOG
                                                        or TOXLINE. If a compound is determined to be of low toxicity,
                                                        it does not need to be considered further for inclusion on  the in-
                                                        dicator parameter list. A compound determined to have hazardous
                                                        toxic properties is further evaluated for properties favoring offsite
                                                        transport.
222
ENDANOERMBNT ASSESSMENT

-------
                                                               Table 1
                                                       Site Specific Indicator List
                WATER SOLUBILITY
 CHEMICAL NAME         at  20°C
Styrene
Phenol
Benzene
                  300 mg/1
                  82 g/1
                  1 ,780 mg/1
VAPOR PRESSURE
   at 20°C


  5 mm Hg
                                      0.2 mm Hg
  76 mm Hg
                                                     BIODECRADABILITY
                                                      BOD   65% ThOD
200 ppm; 57* degraded
within 5 days
Bacterial inhibition
at 6
-------
synergistically  to enhance the magnitude of the toxic response,
or they may act  to detoxify the mixture either by inhibitory or
chemical means.  Exposure assessments are usually conducted on
one chemical at a time, relative to sometimes sketchy or sparse
literature values.  Bioassays can be used to monitor the actual tox-
icity of the  waste mixture as it occurs at various points within a
site.
  Bioassays can also address those components which the analyti-
cal chemical survey or  indicator parameter evaluation may have
missed. In this  respect,  bioassays can be  used as a mechanism for
verifying risk estimates  or acting as a fail-safe mechanism as well
as investigating a waste  site for which no information  is available.
A bioassay  program can be designed which addresses the source
and transport components of the risk assessment, and depending
upon the  type of test organisms used, can also function to model
the responses of  a specific receptor. The calculated result of the
bioassay is a probability which can be reported as risk at various
levels of exposure.
   Table 2 presents the results of a battery of bioassays conducted
using  surface and groundwater  collected  from an uncontrolled
waste site. There  was a concern that materials leaving the site with
groundwater were entering a nearby river and causing  harm to
aquatic life. Figure 2 is  a diagram of the  site and the sampling lo-
cations.
   The bioassays utilized were a static acute 96-hour fathead min-
now (Pimephales promelas) assay, a static acute 48-hour Daphnia
magna bioassay and a static 120-hour algal inhibition bioassay us-
ing the green  algae Selenastrum capricornutum. The bioassays
were conducted in accordance with EPA-Level 1 protocols for the
biological evaluation of complex affluents' and guidelines estab-
lished by the New York State Department of Environmental Con-
servation.'
   The bioassay organisms were chosen to represent aquatic life at
three trophic levels.  The green algae is a primary producer which
fixes photosynthetic energy and provides food for higher trophic
organisms. Daphnia magna is a small freshwater crustacean which
grazes on plankton  and algae, while the fathead minnow repre-
sents a predatory  animal high in the aquatic food chain.
  The minnow  and Daphnia bioassays were conducted in triplicate
at five levels of dilution using ten organisms per dilution. Dilution
and control water were collected from the river, upstream of the
                                                                                                     MWS
                                                                                             MW«
                                                             swev
                                                                               RIVER
                                                                                                    )SwT
                                                                                             MW'HOMTORiNC  WELL
                                                                                             sw. SURFACE WATER
                                                                                    Figure 2
                                                                          Diagram of Waste Disposal Area
                                                         area of the landfill. The groundwater samples were collected from
                                                         shallow wells drilled into the uppermost, unconfined aquifer. Bio-
                                                         assays were initiated within 24-hours of sampling. The algal assay
                                                         was conducted  by seeding dilutions of samples in  nutrient solu-
                                                         tion with algal cells from an actively growing culture.
                                                           The results of the bioassay indicate that groundwater collected
                                                         from within the disposal area is acutely toxic to aquatic life. The
                                                         dilution LCX value corresponds to the concentration of sample in
                                                         dilution  water  at which 50f»  of the exposed  test  population is
                                                         killed. Undiluted groundwater from within the disposal site killed
                                                             Table 2
                                                    Results of Aqualk Bloassays
                                                    Fathead Minnow
                                                                            Daphnia

Site
1
2
3
"4
5
6
7
8

Location
Upgradfent Well
Well In Waste Area
Well In Waste Area
Well 200' Downgredlent
Well 200' Downgradlent
Well 1(00' Downgradlent
Downstream Surface Water
Upstream Control
Survival in Dilution
100 > Sample LC50
10/10 >100*
0/10 2
0/10 8
8/10 >100*
7/10 >100*
10/10 >100\
10/10 >100*
10/10 >IOO*
Survival in Dilution
100* Sample LC50
10/10 >IOO*
0/10 »
0/10 7
6/10 >IOC*
8/10 MOO*
10/10 >100*
10/10 >100*
10/10 >100\
SO* Inhibition
Lev.l
None
m
J7%
None
None
None
None
None
                   10/10 • number of survivors/number at start of  test
                  LC_. > concentration of  sample In dilution water  at which  SO* mortality occurs.
224
ENDANGERMENT ASSESSMENT

-------
all of the exposed minnows and Daphnia. The dilution LC50 values
for water from within the site are less than 10%. Groundwaters
with dilution LC?0 values of less than 10% are considered highly
toxic. The algal bioassays of water from sampling sites 2 and 3 were
also determined to be inhibitory to the growth of algae. Inhibition
of the growth of algae by 50% was calculated to occur in a solution
containing 24% of groundwater from site 2 in nutrient medium and
37% at site 3.  No significant inhibition of algal growth was ob-
served at any of the other sites.
  The bioassay results of samples collected at the upgradient site
and most of the downgradient wells did not reveal any major tox-
icity. Wells 4 and 5, situated approximately 200 feet downgradient
to the disposal area, displayed toxicity in the  100% groundwater
samples. The dilution LC50 values of >100% indicate that a 100%
solution of groundwater could not induce a 50% mortality in the
test populations.  Nevertheless, mortality as  high as 40% was re-
corded in Daphnia at site 3.

CONCLUSIONS

  It can be concluded from  the bioassays that wastes in the dis-
posal area contain components which are toxic to aquatic life. The
results indicate that some of the toxic components may be leach-
ing out of the disposal area and transported by the movement of
groundwater. However, samples collected from wells further down-
gradient or  from  surface water collected at the river indicate that
acute impacts are not being experienced offsite.
SUMMARY

  This paper presents two approaches to the generation of data for
use in remedial investigations of uncontrolled  hazardous waste
sites. Both methods are designed to provide information indepen-
dent of analytical chemical determinations and attempt to integrate
the source and transport elements of a risk assessment. A distinc-
tion is made between hazard and risk because of the dangers of
concluding that a risk exists based on a single hazard term, such as
toxicity, without evaluating mechanisms for exposure or the exis-
tence of receptors.
  The site-specific indicator parameter scheme outlined is a prom-
ising technique which focuses on those chemicals which can be the
source of risk to exposed  populations. It allows chemical analyses
to  be geared  toward a  specific set of chemical compounds by
 methods which can be devised  to provide acceptable,  defensible
 levels of analytical precision and accuracy. However, the site spe-
 cific parameter scheme relies heavily on existing information. Its
 use is  limited to those instances where adequate information on
 site operations and  a cooperative responsible  party  is available.
 Its utility dissolves for sites where little information is available on
 the nature of activity or wastes.
   The bioassay technique described is  a rapid,  relatively inex-
 pensive method for evaluating toxicity. It is well suited for the eval-
 uation of complex mixtures. Used in conjunction with an analytical
 survey for site specific indicator parameters, as well as some generic
 assays such as Total Organic Carbon, bioassays can function as a
 double check for  detecting unidentified toxic components. They
 can also function as an indication of site risks.
   The bioassays presented suffer from the inability to provide in-
 formation on chronic toxicity. At the  present time, most aquatic
 bioassays of chronic toxicity  are much more difficult to perform
 than the statis acute bioassay. However, these assays can be carried
 out where  necessary, and simpler techniques are currently being
 developed. The aquatic bioassays also are not suitable for the esti-
 mation of toxicity to  humans. However, bioassays can be devised
 using small mammals as well as microbial or cell culture  assays
 which can be used to estimate risks of injury to humans.
REFERENCES

1.  Handbook  of Environmental  Data on  Organic  Chemicals, Ver-
   schueren, K., VanNostrand Reinhold, New York, 1983.
2.  Dangerous Properties of Industrial Materials, Sax,  N.I., Van Nos-
   trand Reinhold, New York, 5th Edition, 1979.
3.  Handbook of Chemical Property Estimation Methods, Lyman, W.J.,
   Reehl, W.F. and Rosenblatt, D.H., ed., McGraw-Hill Book Company,
   New York, 1982.
4.  Registry or Toxic Effects of Chemical Substances, Tatken, R.L. and
   Lewis, R.J., ed.,  National Institute  for Occupational  Safety  and
   Health, 1981-82.
5.  "Level 1  Environmental Assessment Biological Tests", USEPA-EPA-
   600/8-81-023.
6.  New York Department Environmental Conservation Draft Guidelines
   for the Performance of Aquatic Bioassays, 1983.
                                                                                   ENDANGERMENT ASSESSMENT
                                                          225

-------
THE IMPORTANCE  OF THE ENDANGERMENT ASSESSMENT
                     IN  SUPERFUND FEASIBILITY  STUDIES
                                        ANNE MARIE C. DESMARAIS
                                                 PAUL J. EXNER
                                                 GCA Corporation
                                              Bedford,  Massachusetts
 INTRODUCTION
   The investigation, assessment and eventual solution of problems
 at uncontrolled hazardous waste sites has evolved in the years since
 the passage of CERCLA into a set of prescribed steps based upon
 the National Contingency Plan and leading to cost-effective, feas-
 ible and environmentally sound removal of actual and potential
 threats to human health, welfare and the environment. One of the
 most important steps toward hazardous waste site remedial action
 is the Endangerment Assessment, a relatively new concept in Super-
 fund actions.
   The Endangerment Assessment combines site evaluation, chem-
 ical  fate and  transport evaluation, basic toxicology and risk and
 exposure assessment into a description and quantification of actual
 and potential hazards and risks associated with the site. The En-
 dangerment Assessment  requires a multidisciplinary  effort. The
 lead role is carried by the environmental  lexicologist with close
 cooperation from the hydrogeologist, chemist and engineer. The
 stated purpose of the Assessment is to determine where remedial
 action is required at a site to mitigate actual or potential threats to
 human health, welfare or the environment and to  access the level
 to which site cleanup is required.
   A preliminary  Endangerment Assessment  should be conducted
 at all hazardous waste sites in general, and NPL sites in particular,
 at the initiation of the Remedial Investigation in order to focus the
 data collection efforts. The final Endangerment Assessment, which
 is developed after the  Remedial Investigation is completed, is a
 bridge between the Remedial Investigation and Feasibility  Study
 and establishes the objectives of the latter.
   The Endangerment Assessment  must characterize a hazardous
 waste site thoroughly in terms of contaminant source,  pathways
 and receptors. The source characterization will define the hazards
 (i.e., chemicals) associated with the site; characterizations of path-
 ways and receptors define the exposure. Assessment of both haz-
 ards and exposure leads  to the description and quantification  of
 risks. It  is important that the Endangerment Assessment considers
 both hazards  and exposure to  deliver to the remedial engineer a
 reasonable and reliable representation of the types and degrees of
 threats posed by the site.
  Source characterizations are as varied as  the hazardous waste
 sites themselves. The Endangerment Assessment team must care-
 fully review all data available in order to assess the character of the
 site and the chemicals present. Pathways usually are limited to four
 media: air, water surface, groundwater and soil, but other media,
 i.e.,  fish, game, crops, etc., may be important at certain sites. Re-
ceptors must be identified for each site based  on the pathways and
area demography.
                                                     SOURCE CHARACTERIZATION
                                                       To characterize the source, the Endangerment Assessment util-
                                                     izes information available  through  previous  investigations  to
                                                     describe the  site  location,  appearance,  topography, geology,
                                                     hydrology and history. In addition, it describes the chemical and
                                                     physical properties  of the contaminants detected at  that site in
                                                     terms of the potential for those pollutants to be released from the
                                                     site to the air,  groundwater, surface water or soil and to contribute
                                                     to endangerment of area populations, including sensitive popula-
                                                     tions, wildlife and/or the environment.
                                                       As part of the data review for the  Endangerment Assessment,
                                                     all hazardous substances detected in air, groundwater and surface
                                                     water, soils, sediments and waste at the site are identified. These
                                                     data are more easily managed if they are categorized by the en-
                                                     vironmental media  in which they were  found. The physical and
                                                     chemical properties of the chemicals are reviewed, including molec-
                                                     ular weight, chemical formula,  vapor pressure,  water solubility,
                                                     solubility in other solvents, biologic detection limits (taste and/or
                                                     odor) in air and water,  bioconcentration factors, soil/sediment
                                                     adsorption coefficients, octanol/water partition coefficients, melt-
                                                     ing and boiling points, degradation rates in water, soil and in bio-
                                                     logical media,  analogous compounds on the basis of structure-ac-
                                                     tivity relationships  and  other factors that may affect the  sub-
                                                     stance's behavior in  the environment.
                                                       When several  chemicals of the same type are present at a site in
                                                     similar  concentrations, they  may  be treated together  in the En-
                                                     dangerment Assessment  if they have similar toxicological proper-
                                                     ties. Chemicals with similar structures often behave similarly in the
                                                     environment and have similar toxicities. The table(s) developed as
                                                     part of the description of the chemical and physical properties can
                                                     be used to compare  properties  of chemicals and to assess  their
                                                     potential environmental behavior.

                                                     Toxicology Assessment

                                                       Following identification of chemical  hazards present at a site,
                                                     each chemical or chemical class is assessed to determine its toxicity
                                                     to people and wildlife.
                                                       The profiles are the result  of literature searches to identify rele-
                                                     vant toxicity data, epidemiological studies and clinical case studies.
                                                     Since extrapolations from test animal to population of concern and
                                                     dose (usually high to low) are almost always required, uncertainty
                                                     is introduced.  Topics and data considered in hazard assessment in-
                                                     clude:
                                                     •Pharmacokinetic Properties
                                                       The four pharmacokinetic processes of adsorption, distribution
                                                     (and storage), metabolism (transformation) and  excretion are re-
226
ENDANGERMENT ASSESSMENT

-------
viewed for each chemical.  These processes describe the mechan-
isms by which people may be exposed to a chemical,  and how it
behaves chemically once it is absorbed into a living organism. The
pharmacokinetics of a chemical will influence its toxicity.
•Acute Toxicity
  Information  about human, laboratory animal and  aquatic life
acute toxic effects (LD50's, etc.) of a chemical are presented. This
information enables one to make comparisons between chemicals
in terms of relative toxicity.  Organ systems affected will also be
noted. Human data are reported where available.
•Subacute and Subchronic Toxicity
  Results of short-term toxicity studies are also reported.  These
studies give a better assessment of the target organs of a chemical
than acute studies that generally expose an animal to a single high
dose of a chemical.
•Chronic Toxicity
  Results of long-term studies in laboratory animals are described
along with reports  of long-term human exposures, as in occupa-
tional settings.
•Epidemiology
  Epidemiological  studies of the  chemicals under  review are
studied. These studies provide the best assessment  of potential
human effects of chemicals.
•Clinical Studies and Case Reports
   These studies, although not always as complete as epidemic-
logical studies, provide excellent information on effects of chem-
icals  on people. Clinical studies are controlled exposures of volun-
teers to chemicals, usually at low levels. They provide insight into
organs  affected by exposures and levels of chemicals at  which
effects  occur. Case studies provide information on single cases of
uncontrolled chemical exposure, often to  high levels. These ex-
posures are generally accidental or occupational.
•Mutagenicity
   Chemicals that are mutagenic in bacteria or mammalian cell
culture are often carcinogenic in laboratory animals. This informa-
tion is important in evaluating the effects of hazardous chemicals.
 • Teratogenicity
   Effects of chemicals on unborn organisms is evaluated. Human
data are included where available.
 •Carcinogenicity
   Although toxic (noncarcinogenic) effects are critically important
in  identifying  chemical hazards, carcinogenic  effects generally
cause the most concern among the population  at large. Carcin-
 ogenicity data  are reviewed thoroughly to assess potential carcin-
 ogenic risks resulting from a site.

PATHWAY AND RECEPTOR CHARACTERIZATION
   The  preceding exercises permit the identification and charac-
terization of a hazardous waste site and the chemical hazards asso-
ciated with it.  In order to be of use to the engineer coordinating
the feasibility study, the information gathered and assessed con-
cerning the site must be  screened and analyzed to direct remedial
action properly.
   Two fundamental questions must be answered: who is exposed,
and  to how much? The  answers are achieved through pathway
and receptor identification and assessment. The information that
was presented in the physical and chemical properties tables is half
of the basis of the pathway characterization and resulting exposure
assessment.  The other half is the assessment of  physical, topo-
graphical, geological and hydrological features of the site.
   The assessment of physical and chemical properties of substances
found at the site permits evaluation of the potential for each chem-
ical to migrate within, between and among environmental media;
the site characterization identifies these pathways of exposure that
are most critical for the site. Routes of exposure (pathways) are de-
fined by  evaluating each possible migration pathway  in terms of
populations and/or environments near or distant from  the site that
may be affected via that pathway. The following factors that must
be  addressed are:
Demography
  The Endangerment Assessment evaluates the number of people
possibly affected by the site, the distance and direction from the
site in which they live and the source(s) of their drinking water.
The demographic evaluation includes people who live, work, play,
shop or go to school near the site.
Environmental Description
  Because things other than people are also affected by hazardous
waste sites, the Endangerment Assessment looks at ecological hab-
itats near the site. Plants, animals and microorganisms excepted
in the site area are described, with special emphasis on unique eco-
systems and threatened or endangered species. Where specific plant
and animal habitats are not known, assumptions of types of organ-
isms living  in an area can be made based on the type of environ-
ment and the geographic area.
Topography and Hydrography
  A thorough description of surface land  and water features en-
ables an assessment of direction and effects of overland flow and
runoff to be made for a site. These pathways are critical, especially
if surface water is a source of potable water in the area or if surface
water is used for recreation.  Surface water is also a pathway for
exposure to fish which are consumed by people and wildlife and
to crops via irrigation.
Meteorology and Climate
  Knowledge of prevailing meteorological conditions at a site en-
ables a description of populations potentially at risk of exposure
to airborne contaminants. The most important information is pre-
vailing wind direction and average wind speed. If possible, monthly
and/or seasonal wind roses  are obtained or can be generated if
sufficient data are  available. This information  is used to assess
whether the air route is important as a means of exposure in a given
area.
Geology and Soils
  A description of soil and rock types underlying the site is impor-
tant for evaluation of migration or stabilization of chemicals in the
porous media. Fractured bedrock or permeable soils and rock are
often critical  pathways  for  transmission of  contaminants from
hazardous waste sites.
Hydrology
  Hydrological and hydrogeological data provide information to
assess migration of chemicals from the site in groundwater. Ex-
posure through drinking water may be one of the most important
routes of exposure associated with a CERCLA site. Relationships
between surface and groundwater are assessed in order to further
define routes of migration  and  relationships between environ-
mental media.
  This information is used to determine where exposures to con-
taminants from a site may occur. Air, surface water  or ground-
water modeling are often required to assess who the receptors are
and to what levels of chemicals they may potentially be exposed.
The following routes of exposure are commonly evaluated:
Air
  Humans can be exposed to airborne contaminants  due to the
volatilization of chemicals from the soil surface or from the surface
of leachate or bodies of water such as wetland areas or streams.
Another source of airborne exposure is dust generated by wind or
activities on a site. People at risk of exposure to airborne contam-
ination will be those living downwind from points of discharge to
air. Through use of wind rose data, the percentage of time during
which specific populations can be at risk of exposure can be calcu-
lated. Airborne contaminants can also contribute to surface water,
soil or vegetation contamination when they settle out or are washed
out  of the atmosphere. The interrelationships  of air  and other
media are assessed for each site depending upon site-specific char-
acteristics.
Surface Water
  Humans can be exposed  to contaminants  in surface water by
drinking the water or by contact with it. Uses of air surface waters
are explored to assess whether waters near a site are used for drink-
ing or human contact.
                                                                                   ENDANGERMENT ASSESSMENT
                                                          227

-------
Groundwater
  Groundwater used as drinking water can be a route of exposure
at many sites. Uses of groundwater in the vicinity of each site will
be reviewed to determine whether people can be exposed through
this route. Interrelationships between surface and groundwater are
evaluated.
Direct Contact
  Contact with contaminated soils or exposed wastes on a site may
be a significant route of exposure, especially if people or animals
can easily gain access to the site.
Ingestion of Contaminated Fish
  People who eat fish that have been exposed to contaminated
surface water, sediments or organisms can be exposed in this man-
ner. Humans who live far from a site may be exposed if commer-
cial fisheries or widely used sport fishing areas are involved. The
Endangerment Assessment evaluates use of fish from sireams and
lakes downstream from hazardous waste sites.
Ingestion of Contaminated Plants or Animals/Birds
  Wild or domestic animals and birds used as human food become
contaminated if their food or water supplies are contaminated.
People may also be  exposed to food crops  contaminated by  ex-
posure to chemicals  in surface water, groundwater  or air. In all
cases of fish, bird, land animal or crop contamination, evaluation
of uptake and bioconcentration factors is important.
 RISK
   After routes of exposure are identified and evaluated, risk assess-
 ment must be conducted  to further  identify those routes of ex-
 posure that contribute significant threats to human health,  wel-
 fare or the environment. The risk assessment conducted as part of
 the Endangerment Assessment is a screening process to determine
 the goals of remedial action.  The risk assessment integrates ex-
 posure and hazard to identify baseline threats at the site. The engi-
 neer addresses these threats and, in the feasibility study, addresses
 the means by which these threats can be mitigated.
   Risk assessments may be quantitative or qualitative. Quantitative
 assessments can be made for substances and routes of exposure for
 which sufficient information is available. Qualitative assessments
 are made when there is more limited information and when  spe-
 cific effects, such as carcinogenicity, can be related to exposure but
 levels of exposure causing those effects cannot be evaluated.
   In GCA's risk assessment approach, human health criteria and
 evaluations developed by the U.S. Food and  Drug Administration,
 USEPA, World Health Organization and other scientific bodies are
 used to evaluate human health effects from hazardous waste sites.
   Data  are likely to be presented in terms of No-Observable Ad-
 verse Effect Levels (NOAELs) for noncarcinogenic effects; a safety
 factor is applied to these data to derive the  Acceptable  Daily In-
 take (ADI) for humans. Carcinogenic compounds often are not ex-
 amined for other chronic effects, since those other effects generally
 result from high exposures than those that  cause carcinogenic
 effects.  A generally acknowledged  exception  is teratogenicity,
 which cannot be predicted quantitatively.
   Noncarcinogenic effects generally are exhibited only after ex-
 posure has reached a threshold; lower exposure will cause no effect,
 and higher exposure is assumed to always elicit the response. The
 threshold is determined by animal testing; that  value is converted
 to what is essentially considered to be the human threshold, the
 ADI.
   The range of exposures to each pollutant is compared to the  ADI
 for that pollutant if one exists. Exposures exceeding that ADI will
 be assumed to be capable of causing the health effect.
   An exposed individual's probability of developing cancer  from
 ingesting contaminated fish can be predicted for many of the pollu-
 tants of interest.  The  USEPA Carcinogenic Assessment Group
 (CAG)  has developed risk  scores in (mg/kg/day) -'  for  many
 pollutants suspected to be human carcinogens based upon animal
 studies. The  exposures calculated for average and highly exposed
 persons are multiplied by the unit risk scores; the result is that per-
                                                        son's  lifetime probability of developing cancer from exposure via
                                                        that route.
                                                          Risk assessments addressing environmental degradation other
                                                        than adverse human health effects are also possible. Generally, the
                                                        data base and the overall understanding of the exposure and the ex-
                                                        posed  population are not as complete as for  human risk assess-
                                                        ment.  Therefore, reliable quantitative  environmental assessments
                                                        are infrequent. The procedure for evaluating environmental threats
                                                        is similar to  that for assessing human  health effects; the data re-
                                                        viewed and the  endpoints selected  for evaluation are  those  that
                                                        affect wildlife, biota, benthic organisms, etc.
                                                          Human welfare issues are more difficult to evaluate and remed-
                                                        iate. Issues such as decreased  property values, loss of natural re-
                                                        sources, loss of development potential in an area and emotional
                                                        stress associated with living near a hazardous waste site should also
                                                        be considered in the Endangerment Assessment.

                                                        CONCEPTUAL REMEDIAL OPTIONS

                                                          At sites requiring a full Remedial Investigation and Feasibility
                                                        Study, an engineer will rely on the results of a  preliminary En-
                                                        dangerment Assessment, conducted at the initiation of Remedial
                                                        Investigation, to  develop a  detailed scope of work for the field
                                                        effort. The preliminary  assessment will clearly define data needed
                                                        to assess the nature and extent of the problem.  The preliminary ex-
                                                        posure and hazard assessments are  important  to the development
                                                        of conceptual remedial options which may be applicable to the site.
                                                          The preliminary exposure assessment will identify the contami-
                                                        nant pathways posing important threats  to area receptors. This
                                                        assessment will aid the engineer in determining what environmental
                                                        media must be addressed and the general locations for remedial
                                                        actions. The  engineer can then develop a list of conceptual reme-
                                                        dial options such as:
                                                        •Excavate contaminated soils at source, treat and dispose on-site
                                                        •Recover contaminated groundwater,  treat and discharge to  sur-
                                                         face water
                                                        •Provide alternative water supply
                                                        •Provide permanent site security
                                                        • Install a groundwater barrier wall
                                                          For  each conceptual option, a list of preliminary remedial tech-
                                                        nologies can  be developed. For the soil excavation and treatment
                                                        option presented above, preliminary technologies might include:
                                                        •Soil excavation
                                                        •Dust control
                                                        •Biological treatment
                                                        •Incineration
                                                        •Waste stabilization
                                                        •Disposal in on-site RCRA-approved landfill
                                                          The preliminary  hazard assessment will  identify  the types of
                                                        chemicals at  the source and how they have been released.  It will
                                                        also identify  those compounds, due to their mobility and toxicity,
                                                        which  pose the greatest threat to public health and the environ-
                                                        ment and,  therefore, which  must be given highest priority during
                                                        site remediation.  The engineer will  use this information to screen
                                                        the  preliminary  technologies,  in particular, those which address
                                                        treatment (e.g., a strictly heavy metals problem would preclude the
                                                        use of biological treatment).
                                                          The engineer undertaking  the investigation must identify prelim-
                                                        inary technologies before going into the field to collect data. With
                                                        a manageable number of technologies in mind,  the engineer can de-
                                                        sign a sampling and analysis program  to collect engineering data
                                                        necessary to  evaluate the appropriateness  of each. This approach
                                                        reduces the need for costly and time consuming returns to the field
                                                        during the Feasibility Study phase of the project.

                                                        IDENTIFICATION OF REMEDIAL ALTERNATIVES
                                                          The final  Endangerment  Assessment is conducted at the com-
                                                        pletion of the data collection effort. The final exposure and hazard
                                                        assessments can be used by the engineer to refine the list of con-
                                                        ceptual remedial options  and  to further screen remedial technol-
                                                        ogies for applicability to the site problem.
228
ENDANGERMENT ASSESSMENT

-------
  The most important step in the Feasibility Study process is the
establishment of response objectives. Each project will have many
objectives. In particular, the engineer must address a number of
"institutional"  issues such  as compliance with RCRA, TSCA,
OSHA and other Federal, State and local regulations. However, at
the core of each remedial action is the need to mitigate the threat to
public health, welfare and the environment  as  specified  by the
NCP. It is the  Endangerment Assessment  which  defines these
threats and answers the question of how clean is clean as specifical-
ly as the data base will permit.
  The engineer  should  not  always expect an exact set of target
concentration levels for each contaminated media. That level of de-
tail is normally beyond the  capability  of  most Endangerment
Assessments since neither the site data base nor the existing toxi-
cology research base is sufficiently detailed. This is particularly true
for human exposure to noncarcinogens and for environmental ex-
posure. The engineer should typically expect a qualitative analysis
of health and environmental threats and a clear presentation of
their priority. That information can then be used to generate a set
of site specific objectives for the response. Examples include:
•To   the  extent  practicable,  remediation must prevent  direct
 human/animal  contact exposure to the contaminated soils in spe-
 cific areas
•Remediation must preclude the human consumption of ground-
 water containing contaminants above concentrations established
 based on toxicity
•To the extent practicable, remediation should minimize erosion of
 contaminated site soils and runoff to surface water
   Note that the last objectives may mitigate effects on biota, name-
 ly fish, in the stream.  However, if  human consumption of con-
 taminated fish is also an important exposure route, then halting
 fish consumption would be an additional response objective.
   After the response objectives and some site-specific design cri-
 teria have been established, the engineer can combine applicable
remedial technologies  into  specific  remedial  alternatives which
must then be screened to eliminate those that are clearly too costly
for the reduction in threat provided.
EVALUATION OF ALTERNATIVES
  Two important elements of the detailed evaluation of remedial
alternatives are the public health analysis and the environmental
analysis. These are conducted in conjunction with technical, cost
and institutional analyses.
  The purpose of the public health analysis is to  evaluate the re-
duction or, in some cases, increase in threat to public health as a
result of the implementation of each remedial alternative. The
baseline or "no action" threat is that established by the Endanger-
ment Assessment. The analysis can be either quantitative or quali-
tative, depending upon the types of contaminants and/or the extent
of the data base.
  The purpose of the environmental analysis is to  evaluate reduc-
tion  or increase in threat to biota from remedial alternative imple-
mentation. Since the toxicology research base is limited, this analy-
sis will be qualitative. Again, the baseline threat is  that established
by the Endangerment Assessment.
  Ultimately, the selected remedial alternative will be that which is
technically reliable, satisfies institutional criteria (including public
welfare  concerns)  and cost-effectively mitigates threats to public
health and the environment.  At any hazardous waste site where
cleanup funds are limited (there are very few outside of this cate-
gory), remedial action must be prioritized and funds allocated to
mitigate threat to public health first and threat to the environment
second.  It is clear that the engineer must rely heavily on the ex-
pertise of the environmental toxicologist and must  thoroughly
understand, and be able to contribute to, the site remediation tool
known as the Endangerment Assessment.
                                                                                  ENDANGERMENT ASSESSMENT
                                                          229

-------
 HEALTH  RISK  ASSESSMENTS  FOR  CONTAMINATED  SOILS

                                                  KARL L. FORD
                                                  PAUL GURBA
                                         Ecology and Environment,  Inc.
                                                 Denver, Colorado
 INTRODUCTION
   Hazardous waste contaminants are known to partition between
 environmental media based on their physical and chemical proper-
                                                     and those substances which at low doses pose a small but  finite
                                                     cancer risk. Separate methods are presented for each type of con-
                                                     taminant.
ties.1' 2 Both metals and organic contaminants tend to sorb to a     Threshold Contaminants
greater or lesser degree onto organic matter and clay particles in the
soil and sediment.1'3 Consequently, soils serve as a sink for many
contaminants, particularly cationic metals and the organic com-
pounds with low water solubility and high sorption coefficients.
  Of the various pathways of exposure associated with hazardous
waste sites, the soil pathways are among the  most important but
least understood. Exposure from contaminated soils can arise from
inhalation of dust, incidental ingestion via soiled hands, bioaccum-
ulation by crops and livestock and leaching into surface or ground-
waters used for drinking. Experience has indicated that exposure
via incidental soil ingestion may be the critical pathway at many
sites.
  One of the difficulties in evaluating health risk from contam-
inated  soils  is the general  lack  of health criteria  for the  soil
medium. Unlike the air and  water media, health specialists have no
convenient criteria to  serve as a basis for comparison. To be sure,
the USEPA has established  criteria  for land application of sewage
sludge to land,4 and agronomists have also established general guid-
ance on  tolerable levels of soil contamination  for agriculture.1
However, these criteria are more appropriate in evaluating soil bio-
accumulation or crop and animal toxicity potential.
  A simple, reliable method of deriving soil criteria based on in-
cidental soil ingestion is needed for determining the health risk of
contaminated soils in or near residential areas. Public health
specialists have many years of experience in evaluating the relation-
ship between the  pica habit  of soil ingestion and urban childhood
lead poisoning.'  Similar studies have  indicated  that, for heavy
metal exposure around smelters, the soil ingestion pathway may be
the most critical exposure pathway.'
  Until recently,  a troubling aspect of this approach was that not
enough was known about the amount of soil  actually ingested by
children. More research is required  on this subject. However, the
Centers for Disease Control of the U.S. Public Health Service has
established soil ingestion figures for  various age groups of the pop-
ulation (Table 1).' In this paper, the author discusses  a method de-
vised to utilize these soil ingestion data and certain health criteria
to calculate health-related soil criteria.
METHODS
  Two separate types  of contaminants are commonly encountered
at hazardous  waste sites: those substances which at small, sub-
threshold doses are not thought to cause any chronic  health effect,
                                                       A health criterion should be designed to protect the most sensi-
                                                     tive fraction of the population. Table 1 shows that the age group
                                                     1.5  to 3.5  years consumes  a disproportionate amount of soil.
                                                     Given their low body weight compared to an adult, in addition to
                                                     their soil ingestion habits, children in this age group are clearly at
                                                     greatest risk of exposure from  contaminated soils  at hazardous
                                                     waste sites. A daily soil ingestion figure of 10 g/day appears to be
                                                     the proper level to consider when evaluating worst case exposure.
                                                       The most useful health criterion for threshold contaminants is the
                                                     allowable daily  intake (ADI). This  intake,  usually expressed in
                                                     terms of mg/day, was originally devised by the FAO/WHO Expert
                                                     Committee on Food Additives.' It was intended to be the adult in-
                                                     take at which it was believed no lifetime health effects would occur;
                                                     however, it was not intended as a guarantee of absolute safety.
                                                       The ADI may be modified by body weight  and daily soil intake
                                                     to calculate the soil criteria (SC). A 10  kg bodyweight and a 10 g/
                                                     day soil intake are used to provide protection for the most exposed
                                                     group. The method of calculation is as follows:
                                                      SC (ug/g) = ADI (mg/day) x lOOOg/kg/SI (g/day) x BW
(1)
                                                      where SI is  the soil ingestion (10 g/day) and BW is bodyweight
                                                      adjustment (10 kg/70 kg). Using phenol as an example, the ADI is
                                                      7.0 mg/day,'0 the soil criteria derived according to this method is
                                                      100 mg/kg (ppm).
                                                      Carcinogens

                                                        The child soil ingestion figure of 10 g/day is not appropriate
                                                      for exposure to carcinogens since it is a figure assigned for a very
                                                      limited age group and does not reflect the lifetime soil ingestion
                                                      exposure. Cancer risk is considered to be cumulative over a life-
                                                      time, hence  soil exposure should also be  evaluated in cumulative
                                                      fashion. A lifetime average soil ingestion (LASI) of 0.006 g/kg/day
                                                      was calculated from the data in Table 1. This figure was obtained
                                                      by summing the lifetime soil ingestion and dividing by 70 kg body-
                                                      weight and by the days in a 70 yr lifetime.  A slightly higher soil
                                                      ingestion figure could be obtained by using a lifetime average body-
                                                      weight.
                                                        Since lifetime exposure is the concern  with carcinogens, some
                                                      allowance should be made for the degradation of the contaminant
                                                      in the soil over the human lifetime. The environmental half-life of
                                                      the contaminant may be used to account for the fate processes
230
ENDANGERMENT ASSESSMENT

-------
                            Table 1
     Soil Ingestion by Age Group, Calculation of Lifetime Average
                      Soil Ingestion (LASI)

Age Group
0-9 mos
9-18 mos
1. 5-3.5 yrs
3.5-5yrs
5-18 yrs
18 yrs

Years
0.75
0.75
2.0
1.5
13.0
52.0

Ingestion (g/day)'
0
1
10
1
0.1
0.1
Sum of
Soil Ingestion (g)
0
274
7300
548
475
1898
                                            10495 g/lifetime
                      lifetime
10495 g/lifetime x 70 kg x 365 days x 70 yrs =  0.006 g/kg/d (LASI)
such as biodegradation,  hydrolysis and photolysis. Of these pro-
cesses, biodegradation data may be the most available." A half-
life adjustment factor of t/2/70 is applied to the calculation of the
soil criteria, where t/2 is half-life in years  and 70 is the approx-
imate human lifespan in years.  Half-life data generally assume
linear decay. No half-life correction is applied to metals because
they are resistant to degradation.
  The health criteria may be obtained from slope estimates (also
called  unit carcinogenic risks, UCRs) published  by the USEPA
Carcinogen Assessment Group.12 These data are expressed as an ex-
cess cancer risk from a lifetime of ingestion of 1 mg/kg/day of a
carcinogen. By selecting an appropriate risk level, a figure loosely
termed the "lifetime allowable daily intake"  (LADI) may be cal-
culated:
LADI (mg/kg/day)  = Risk/UCR
(2)
   For example, the UCR for dieldrin is 30.4/mg/kg/day. At a risk
level of 1x10-6 (1 excess case in 1,000,000 exposed), the LADI is
calculated to be 3.3x10-8 mg/kg/day.  The soil criterion is then
calculated as follows:
 SC = LADI x lOOOg/kg/LASI x t/2/70
(3)
 where SC =  soil criterion (mg/kg), LADI =  lifetime  allowable
 daily intake (mg/kg/day), LASI = lifetime average soil  intake
 (g/kg/day) and t/2/70 is  the half-life correction factor. For ex-
 ample, the soil criterion for dieldrin, with a half-life of 0.14 yr, is
 calculated as follows:
 SC = 3.3x10-8x1000/0.006x0.14/70
 SC = 2.75 mg/kg (ppm)
(4)
 Using DDT, with a UCR of 8.42, the LADI is 1.19x10'' and the
 half-life is 14.6 yr. The soil criterion is calculated to be 0.1 mg/kg.
 Much of the difference is due to the half-life correction factor. As
 this term approaches unity, the soil criterion  decreases. Arsenic,
 with a UCR of 15, a LADI of 6.67x10"' mg/Kg/day and no half-
 life correction, has a soil criterion of 0.01 mg/kg.

 DISCUSSION
   With a global median soil concentration of 6 ppm,13 arsenic is
 an example of a soil criterion derived by this procedure that is less
 than natural background levels. While there is no assurance that
 background levels are not contributing to local cancer incidence, it
 does not seem  reasonable to establish a criterion that is more
 stringent than natural  background  levels. Therefore, the recom-
 mended guidance for  naturally occurring metals is to accept the
 calculated criterion or  background concentration,  whichever is
 greater.
         The soil criterion derived for carcinogens is based on some level
       of "acceptable" cancer risk. While no consensus has emerged on
       an acceptable level of risk, a risk of 1x10"« is useful for a first
       approximation of a soil criterion.
         Some doubt exists about the applicability of the half-life correc-
       tion factor to threshold contaminants. The soil ingestion figure of
       10 g/day is based on an exposure of only 2 yr (Table 1), and it is be-
       lieved that degradation may not be substantial enough in that time
       period to warrant its use. Furthermore, the ADI was not originally
       intended for substances such as heavy metals which exert a cumu-
       lative effect in the body and are not subject to decomposition.
         Other  soil  exposure pathways, including inhalation of wind-
       blown dust and crop and livestock bioaccumulation, are not direct-
       ly addressed by this  procedure.  Limited experience with the ex-
       tremely insoluble  1, 3, 7,  8 tetrachlorodibenzo-p-dioxin  suggests
       that the  exposure and cancer risk from soil ingestion exceeds ex-
       posure from food chain bioaccumulation.8 In addition, several in-
       vestigators studying lead exposure around smelter sites  have re-
       ported that soil ingestion leads to greater exposure than dust in-
       halation.7 Hence, it appears that establishing a soil criterion based
       on ingestion may provide an adequate margin of safety for other
       soil-associated exposure pathways.
         The soil  criteria derived by this procedure appear to provide
       health-based criteria that may be used to assess the health risk of
       soil exposure by direct contact and may also provide an adequate
       margin of safety for other soil pathways. These criteria may be used
       as a first approximation of cleanup levels for remedial action plan-
       ning at hazardous waste sites. Other considerations such as future
       land  use and other soil exposure pathways may  influence final
       criteria selection.
REFERENCES

 1. Kenaga,  E.E.  and Goring, C.A., "Relationship Between  Water
   Solubility, Soil Sorption, Octanol-Water Partitioning and Concen-
   tration of chemicals  in the Biota,"  Aquatic Toxicology, ASTM
   STP707,  Eaton, J.G., Parish,  P.R., Hendrichs,  A.C., eds.,  1980,
   78-115.
 2. Neely, W.B., "Organizing data for  environmental  fate  studies,"
   Environ.  Chem. andToxicol., 1, 1982, 259-265.
 3. Farrah, H., Hatton, D. and Pickering, W.F., "The affinity of metal
   ions for clay surfaces," Chemical Geology, 28, 1980, 55-68.
 4. USEPA,  Process Design Manual for Land  Treatment of Municipal
    Wastewater, Washington, D.C., 1981.
 5. Kabata-Pendias, A. and Pendias, H., Trade Elements in Soils and
   Plants, CRC Press, Boca Raton, FL, 1984, 9-11.
 6. Mielke, H.W., Blake, B., Burroughs, S.  and Hassinger, N., "Urban
   lead levels in  Minneapolis:  the case of the  Hmong children," Env.
   Rev. 34, 1984, 64-76.
 7. Roels, H., Buchet, J., Lauwreys, R., Bruaux, P., Elaeys-Thoreau,
   F., LaFontaine, A. and Verduyn, G., "Exposure to lead by the oral
   and pulmonary routes of children living  in the vicinity of a primary
   lead smelter," Env. Res. 22,  1980, 81-94.
 8. Kimbrough, R.D., Falk, H., Stehr, P. and Fries, G., "Health Im-
   plications of 2, 3, 7, 8 TCDD contamination of residential  soil,"
   submitted to the J.  ofToxicol. and Environ. Health.
 9. FAO/WHO Expert Committee  on Food Additives, Evaluation of the
    toxicity of a number of antimicrobials and antioxidants, World Health
   Organization Technical Report Series. Report 229. Geneva,  1962.
10. USEPA,  Environmental Criteria and Assessment Office,  Draft Sum-
   mary of Published Acceptable Daily Allowances for USEPA Priority
   Pollutants. 1984.
11. Tabak, H.H., Quave, S.A., Mahni, C.I. and Barth, E.F., "Biode-
   gradability  studies with  organic priority  pollutants compounds,"
   JWPCF53, 1981,1503-1518.
12. USEPA,  Environmental Criteria and Assessment Office. Draft Health
   Assessment Document for Chromium. 1983, 7-79 to 7-81.
13. Bowen, H.J., Environmental Chemistry of the Elements. Academic
    Press, New York, NY, 1979, 60-61.
                                                                                     ENDANGERMENT ASSESSMENT
                                                                   231

-------
                       PUBLIC  HEALTH  SIGNIFICANCE  OF
                                HAZARDOUS WASTE SITES

                                             ROBERT  L. KAY, JR.
                                            CHESTER  L. TATE, JR.
                                  Department of Health and Human Services
                                           Centers for Disease Control
                                                 Atlanta, Georgia
CENTERS FOR DISEASE CONTROL'S
INVOLVEMENT
  With the passage of CERCLA, funds have become available to
identify, investigate and cleanup abandoned waste sites containing
hazardous substances, pollutants or contaminants and to provide
immediate response to environmental emergencies involving hazar-
dous substances. This Act provides a national response mechanism
to protect public health,  welfare  and  the environment from any
releases or substantial threat of releases of any pollutant or con-
taminant that may pose an imminent and substantial danger.
  CERCLA  defines pollutants  or contaminants  as  elements,
substances, compounds  or  mixtures, including  disease-causing
agents, which, after release  into  the environment and upon ex-
posure,  will  or could  cause death, disease, behavioral abnor-
malities, cancer, genetic mutation, physiological malfunctions (in-
cluding  reproductive  pathologies)  or physical  deformities  in
organisms or their offspring.  The definition does not include
petroleum or natural and synthetic gas.
  The role of the Department of Health and Human Services
(HHS) under CERCLA is defined in both Executive Order  12316,
"Responses to Environmental Damage," and the National Oil and
Hazardous Contingency Plan [Title 40 CFR 300]. The President
has delegated all authorities  under Section  104 [b] of  CERCLA
relating to "illness, disease, or complaints thereof to the Secretary
of DHHS, who has further delegated these authorities to the Public
Health  Service/Centers   for  Disease  Control   (PHS/CDC).
Whenever the PHS "has reason to believe that....illness, disease, or
complaints thereof may be attributable to exposure to a hazardous
substance, pollutant, or contaminant, and that a release may have
occurred  or may be occurring,"  investigations and information
gathering as determined necessary may be undertaken "to identify
the existence and extent of the release or threat thereof,  the source
and  nature of the hazardous substances,  pollutants, or con-
taminants involved and  the  extent  of danger to  the  public
health...." Because of  CERCLA and  an agreement between the
USEPA and PHS/CDC, the question of whether a site poses an im-
minent and substantial danger or threat to local  public health is
assessed for each site.  The USEPA uses  public health assessment
and evaluation to help determine the priority that  should be given
to a particular site for remedial action.  A site may require both im-
mediate  and  long  term remedial measures  to safeguard  public
health.
  CERCLA further stipulates that the Administrator of the Agen-
cy for Toxic Substances and Disease Registry (ATSDR), James O.
Mason, M.D., who is also the Director of CDC, shall "in cases of
public health emergencies caused  or believed to be caused by ex-
                                                    posure to toxic substances, provide medical care and testing to ex-
                                                    posed individuals, including but not limited to tissue sampling,
                                                    chromosomal  testing,  epidemiological  studies,  or  any other
                                                    assistance appropriate under the circumstances...." [Section 104 [i]
                                                    of CERCLA]. CERCLA requires  the newly created ATSDR to
                                                    establish and maintain a:
                                                    •National registry of severe diseases and illnesses and of persons
                                                     exposed to toxic substances
                                                    •National inventory  of research information on health effects of
                                                     toxic substances
                                                    •National list of all areas closed to  the public or  otherwise re-
                                                     stricted in use because of contamination by a toxic substance
                                                      ATSDR also conducts periodic surveys and screening programs
                                                    to determine relationships between toxic substance exposures and
                                                    illnesses. Its medical epidemiologists and laboratory scientists are
                                                    involved in a number of these screening and health effect studies at
                                                    several Superfund sites. Some of these studues are outlined below.
                                                     STUDY
                                                     1. Battle Creek. MI,
                                                       Health Study

                                                     2. Hollywood Dump
                                                       Site, Memphis, TN
                                                     3. Pocono Summit
                                                       Site, PA
                                                     4. McKin Dump Site,
                                                       ME
                                                    5. Dioxin, MO
                                                     6. Oak Ridge. TN
                                                     7. PCB Study
                                                       Bloominglon, IN
                                                     8. Childhood Lead
                                                       Study, ID
SYNOPSIS/STATUS
Study to assess health effects of low levels
of volatile organic compounds [VOC] in
drinking water
Study to assess exposures and health effects
from insecticides
Possibility of conducting a case-control
study of eosinophilic granuloma
Proposed study to assess health effects of
contaminated drinking water
Health study for Quail Run; Adipose
Tissue Study; and Reproductive Outcomes
Study are awaiting final approval;
preparations have begun
Study to assess mercury uptake through
fish consumption and exposure to
contaminated soil
Study to assess PCBs

Study to assess blood lead levels and lead
levels in soils, house dust and vegetables
due to a smelter; findings are indicating
that increased blood lead levels are
associated with high soil lead levels
 232
ENDANOERMENT ASSESSMENT

-------
9. Childhood Lead
   Study, East Helena,
   MT
10.  California
    Cooperative
    Agreement
11.  Tacoma Smelter
    Study, WA
Study to assess levels and environmental
exposure pathways of lead, arsenic and
cadmium; also analyzing 3-day composite
stool samples for silicon, titanium and
aluminum in order to estimate infant's
dirt and dust ingestion
Agreement to fund a number of studies in
California, to include the BKK landfill and
Silicon Valley groundwater contamination
Study to examine environmental exposures
to arsenic and subclinical health effects
ITEMS OF SIGNIFICANCE

  To evaluate the public health significance of hazardous waste
sites, State and local public health agencies, Federal agencies and
environmental  consulting  firms involved  in  the  assessment of
hazardous waste sites require more resources to understand and
determine the:
•Primary contaminants of concern
•Potential environmental pathways for exposure
•Potential health impacts from exposure
•Objectives of a public health evaluation
•Information necessary to adequately study the health implication
 of a hazardous waste site
•Prioritization  of hazardous waste sites to assign environmental
 testing, remedial measures or health studies to a site because of the
 potential or known danger to public health
  Effective decisions to protect public health can only be made if
sufficient information is available on each hazardous waste site and
the potentially affected local populations and environment. As one
gains information about  the  potential health effects of  toxic
substances and the relationship between exposure and these health
effects along with a better understanding of the transport and ex-
posure  pathways, public health evaluations of hazardous waste
sites will improve.

PRIMARY CONTAMINANTS OF CONCERN

  In conducting a public health evaluation of a hazardous waste
site, scientists need to identify what contaminants of concern exist
on the site. To classify whether a substance at a waste site is a con-
taminant of concern, scientists first need to determine if it meets
CERCLA's definition  of a pollutant or contaminant. Next, they
need to find out about the occurrence of releases and the toxicity of
the  released  substance.  The quantity,  concentration, reactivity,
compatibility, environmental persistence and transport potential of
the substance are other important factors needing consideration to
determine whether a substance at a site is a contaminant of con-
cern.

Occurrence of Releases

  These contaminants  may include substances  from uncontained
spills, releases of a substance and substances that are  likely to be
spilled or released because of the condition of their containment
vessels or structures. If the substance were adequately contained
and  environmental transport and  exposure were not likely, the
substance would not be one of immediate concern.

Toxicity

  The kind of environmental toxicities  that a substance poses to
people and other organisms needs to  be considered. For many
substances, acute toxicity data are available; additionally, longer
term toxicity data may also be available. Toxicologic information
about  the  oncogenic,  teratogenic  and mutagenic effects of  a
substance known to be or suspected of being a genotoxin may be
available. Dose-response data about a substance can provide the
toxicologic information needed to describe the likely occurrence of
an acute or chronic effect or, in the case of carcinogens, a proba-
bility of risk that a cancer death or other disease may occur over a
lifetime of exposure.
  Toxicity ratings and/or data are provided by Sax1 and the Na-
tional Fire Protection Association.2 Toxicity data are also available
from the HHS's National Toxicology Program agencies (including
the  National  Cancer  Institute,3  the National  Institute of En-
vironmental Sciences, the Food and Drug Administration, the Na-
tional Institute for Occupational Safety and Health,4-5 the National
Library of Medicine,' USEPA's Office  of Health Research,7
USEPA's Office of Water Regulations and Standards,' USEPA's
Office of Toxics Integration and  the  National  Academy of
Sciences/National Research Council.'
Quantities and Concentrations of the Substance
  If the quantities and concentrations of the substance at the point
of release are sufficient to cause harmful effects to man or food
chain organisms, the substance would be considered a contaminant
of concern. For substances that are  highly toxic, even very small
quantities and concentrations may pose an unacceptable risk to
public health.  If the substance can be, or is being, transported from
the point of release, the concentrations represented at the sampled
site  of transport and  the degree to which these concentrations
adversely  affect public health would help determine whether the
substance is a primary contaminant of concern. These environmen-
tal  concentrations may be compared with applicable standards,
criteria and guidelines to assess their acceptability in terms of public
health, welfare and the environment.
Environmental Persistence and Stability of the Substance
  The persistence and stability of the substance at the waste site is
an important consideration in determining whether the substance
would be considered a contaminant of concern. A  number of
physical, chemical and biological processes may be important in af-
fecting the degradation and/or distribution of a substance  in the
environment. In many cases, the fate processes are predictable and
can be used to assess the short term and long term significance of
the contaminant to public health.10 These processes can degrade a
contaminant into a relatively harmless state or product at  a rate
which reduces or eliminates the initial hazard posed on or off the
site. Depending on the substance, the processes can also create or
transform it into a more harmful state or product.
  Highly persistent substances with the ability to bioaccumulate in
the food chain are often classified as contaminants of concern, par-
ticularly if environmental transport is demonstrated. Even  if the
waste site contains low levels of these contaminants, their ability to
persist in the environment and bioaccumulate in the food chain can
cause them to exceed permissible levels  in water and food over a
period of time. The  USEPA's National Oil  and  Hazardous
Substances Contingency Plan11 gives persistence ratings for many
organic compounds.
  The stability of a substance also depends on its reactivity, its
compatibility with other substances in the waste site and its mobil-
ity  through different environmental media.  Substances can be
highly reactive when subjected to certain physical conditions and
when mixed with incompatible materials. Sudden releases of con-
centrated  air pollutants may result. The NFPA assigns reactivity
ratings to  a number of common chemicals.2  Chemical incom-
patibility can be determined by examining Table 12, "Incompatible
Materials," in the National Oil and  Hazardous  Substances  Con-
tingency Plan.11
  The mobility or transport of a substance depends on a variety of
chemical, physical and biological processes. Aquatic fate processes
depend upon several physiochemical factors for the substance, in-
cluding vapor  pressure, octanol/water partition  coefficient, lipid
solubility,  aqueous  solubility  and  local  environmental  condi-
tions—pH, sunlight flux, ionic composition  and strength of the
water, redox  potential,  organism population and temperature.
Chemical  speciation, photolysis, oxidation, hydrolysis, volatiliza-
tion, sorption, bioaccumulation and biotransformation are fate
processes that  can affect a substance's mobility. The USEPA has
classified these aquatic fate processes for each of the  129 priority
pollutants.12 Examining the probable fate processes for the con-
taminants will help determine which contaminants on the site are of
concern because of their predicted environmental behavior.
                                                                                  ENDANGERMENT ASSESSMENT
                                                                                                   233

-------
Fire and Explosive Hazards
   If a waste site contains substances that are highly reactive, in-
compatible with other substances on the site or easily ignited, the
potential safety hazards of fire and explosion to surveillance and
monitoring teams and to cleanup crews and local populations must
be evaluated before the site is disturbed in any way. The NFPA has
classified ignitability, incompatibility  and reactivity levels.1 The
U.S. Department of Transportation  has developed evacuation
tables" for different compounds for fire and explosive hazards and
downwind air pollution hazards.
 POTENTIAL ENVIRONMENTAL
 PATHWAYS FOR EXPOSURE
   Human exposure to the primary contaminants of concern usually
 depends upon two pathways:  transport pathways and  exposure
 pathways. Transport pathways are the displacement of the con-
 taminant from a source of contamination to a receptor site. Ex-
 posure pathways are avenues by which humans may absorb or react
 to contaminants through contact, ingestion or inhalation.
   Even though contaminants can be transported  from a source of
 contamination through a  variety of different processes, health
 authorities  need to determine  and  rank the  significance of the
 transport routes in  causing human exposure and  endangering
 public health. Mitigation efforts should be concentrated on these
 routes until the imminent hazard from known or potential exposure
 is eliminated.

 Transport Pathways

   Hazardous materials can be  transported from a waste or spill
 through several environmental pathways. Water, air, soil and sedi-
 ment  are the primary  media  of  physical  transport,  whereas
 organisms in the food chain  (both plant and  animal) are the
 primary media for  biological  transport (Table  1). CDC  has
 developed a screening procedure" for identifying the transport
 pathways and the factors associated with these pathways that in-
                                                         crease the potential risk of exposure. When transport
                                                         cause human exposure to hazardous substances, the public
                                                         significance of the transport pathway depends upon  the duration
                                                         and degree of exposure, the health  hazard posed by the contami-
                                                         nant of concern and the controllability of the transport or exposure
                                                         pathway.
                                                           Site-specific  information about environmental pathways is not
                                                         always  available.  Therefore,  environmental   surveillance  and
                                                         monitoring of the site becomes necessary. Complete characteriza-
                                                         tion of a site requires extensive field work and may be very expen-
                                                         sive, yet still not provide conclusive  evidence; the time  required for
                                                         study can delay implementation of  interim measures to safeguard
                                                         public health. Because of this, preliminary estimates of the concen-
                                                         tration and the fate of  the chemical in the local environment are
                                                         often used to evaluate the hazards of the site, the possible duration
                                                         of past exposures, the expected distribution and rate of contamina-
                                                         tion and the expected worst-case concentrations. These preliminary
                                                         estimates  can also be  useful in determining the type and extent of
                                                         additional environmental sampling  required and the need  for im-
                                                         mediate protective measures.
                                                           Computer-based  emergency  response  programs  have  been
                                                         developed and are now available to  serve as models for the disper-
                                                         sion of  spills and contaminants in  the environment.  These com-
                                                         puterized  modeling programs can simulate the  behavior of toxic
                                                         clouds and plumes in both air and water and incorporate chemical
                                                         source characteristics, hydrological  and meteorological data and
                                                         site-specific characteristics.
                                                           To estimate the fate of toxic substances in the aquatic environ-
                                                         ment, the USEPA, with the help of contractors,  has developed
                                                         screening  procedures"-16 for assessing the fate of toxic substances
                                                         in both  surface and subsurface water. Although these procedures
                                                         have been verified with  field data, users need to  become aware of
                                                         the assumptions,  potential  errors  and limitations associated with
                                                         them. The goal of these screening methodologies is to determine
                                                         and identify, with a minimum of effort, whether either existing or
                                                         projected  loading rates  from toxic  pollutants are likely to reach
                                                         hazardous levels  in water resources. Aquatic fate  predictions  of
 Transport
 Pathways
                                                            Table 1
                                                 Transport and Exposure Pathways
                                          Transport Media or Process
 Contaminants
   In:
         Occupational
         Mitigation
         Measures,
         Cleanup	
Drinking
Water
Swimming/
Bathing
Land/
Recreation
Air Quality
                                                                       Indoor Outdoor
                   Contaminated Pood*
                                            Explosion
                                            Fire
 Ground water

 Surface water

 Soil and
 Sediment

 Dust
 (Windborne
 or  Man-
 Disturbed)

 Rainfall  or
 Fallout

 Leaky  Drums
          Inh.C

          Inh.C


          Inh.C
 Ing

 Ing


 Ing
         Inh.C.Ing     Ing


         Ing            Ing

         Inh.C.Ing     Ing
C.Ing.Inh

C,Ing,Inh


C,Ing,Inh     C.Ing.Inh
                                      Inh
                       Inh
                                 Ing     Ing    Ing
                          Ing,Inh     Inh, Ing  Inh,Ing
                                             C.Ing
                          C.Ing.Inh
                                            SH

                                            SH
                                                                       Inh,Ing   Ing    Ing     Ing
                                                    Ing     Ing
                                   Inh, Ing   Ing    Ing     Ing

                                   Inh       Ing    Ing     Ing      SH
Other
Contaminant
Structures Inh, C, Ing C.Ing C.Ing.Inh


Inh Ing Ing SH
Probable Human Expoiure Palhwayi:
Safely Hazard - SH; Ingrallon - Ing; Skin and eye contact - C; Inhalation . Inh
                                                      •Food chain organium contaminated through: direct contact; Ingestion of contaminated ore nimu
                                                      soil or water; Inhalation; habitation in contaminated water; and plant uptake of contaminants^'
234
ENDANGERMENT ASSESSMENT

-------
pollutants involve the delineation of the physical compartment of
the environment within which the transport processes act and the
identification of the pertinent fate processes:15

•Loading              —Inflow rates of toxicants (from discharge and
                       from atmospheric, land runoff and sediment
                       depositions)
•Speciation             —Acid-base Equilibria (pH)
                      —Sorption
•Transport             —Precipitation/Dissolution
                      —Advection (hydraulic flow)
                      —Volatilization
                      —Sedimentation
^Transformation        —Biodegradation
                      —Photolysis
                      —Hydrolysis
                      —Reduction/Oxidation
JBioaccumulation       —Bioconcentration
                      —Biomagnification
  To simplify the screening procedures and to minimize the effort
that may be expended with a complete analysis of the fate of a toxi-
cant, the USEPA" recommends that one should first assume that
the  pollutant behaves conservatively;  that is, that it  does not
undergo reaction.  This approach requires only data on  pollutant
loads and hydrological parameters rather than environmental data
or rate constants. Since pollutant decay and removal processes are
neglected, this approach will yield the highest possible estimate of
pollutant levels.
  If initial predictions are higher than  a standard, a more refined
approach,  including the fate processes, more information on en-
vironmental and chemical constants12-13-15  and more complicated
equations, would be necessary. If initial predictions are lower than
a standard, a problem is unlikely. However, there are drawbacks to
the conservative approach because it may neglect accumulations of
the pollutant in other environmental compartments (that is, bedded
sediments and any exposure that may result from these other com-
partments.
Exposure Pathways
  If transport pathways or media are contaminated in excess of ap-
plicable standards,  criteria or guidelines, adverse human exposure
through ingestion,  inhalation or  contact could   occur.  When
populations may be subject to adverse exposure, immediate efforts
would be needed to identify necessary remedial measures. Possible
transport pathways and media that could lead to human exposure
are shown in Table 1.
POTENTIAL HEALTH IMPACTS
FROM EXPOSURE
  Without  conducting a  health study  of the biological levels or
health effects of populations with the greatest evidence or risk of
exposure, researchers cannot  obtain precise information on ex-
posures and the resulting health effects.  Potential risk and the
magnitude of exposures can be estimated, however, by using risk
assessments to approximate the public health significance of the ex-
posure. Decision makers can use these estimates, along with infor-
mation from  relevant Federal and State guidelines, criteria and
standards for hazardous substances in the various transport media,
to assess the significance of the potential health impact and the im-
minency of the health hazard  and to decide on  the  need for
remedial actions.  CDC has developed a  screening manual,14 A
System for Prevention, Assessment and Control of Exposures and
Health Effects from Hazardous Sites (S.P.A.C.E. for Health), to
help States prevent, assess and control exposures and health effects
from hazardous substances at these sites. These screening pro-
cedures will help State and local health agencies identify sites that
need health studies. The purpose of the health studies are:
•To further assess the risk to human health posed by a given site so
  that appropriate preventive actions and public health  measures
 can be taken as needed and appropriate medical advice can be
 provided
•To advance scientific knowledge about the persistence and effects
 of hazardous substances in the body so that risk in other situa-
 tions can be assessed
  At many hazardous sites, health studies may not yield significant
or meaningful results or may not be feasible. Therefore, it is par-
ticularly  important to recognize  circumstances  in  which health
studies are likely to be useful and feasible.

OBJECTIVES OF A PUBLIC
HEALTH EVALUATION
  The primary objective of a public health evaluation is to address
the questions posed by the requester. In addressing these questions,
the public health official may need to appraise, investigate or iden-
tify the:
•Significance of each known environmental transport/media path-
 way (for example,  groundwater,  surface water, air, soil, sedi-
 ment and food chain) as a cause of human  exposure
•Significance of each exposure  pathway  (for example, ingestion,
 inhalation and body contact)
•Health effects that  have  already occurred or could occur from
 known exposure to  the contaminants of concern
•Imminency of the health hazard via environmental exposure
•Imminency of the potential human health hazard via environ-
 mental transport
•The need for immediate actions to prevent, limit or mitigate an
 emergency situation because of the imminent risk or danger to
 public health
•The need for long  term actions  to prevent, limit or mitigate a
 potentially dangerous  situation in the future
•The need for conducting epidemiological studies and for provid-
 ing medical care and testing to persons exposed to toxic sub-
 stances as a result of a public health emergency
•The health disorders  and  diseases that  may be associated with
 exposure to a particular toxic  substance; health disorders could
 include  behavioral  abnormalities, cancers, genetic  mutations,
 physiological malfunctions, reproductive malfunctions and physi-
 cal deformations including birth defects
•The need for  reviewing and comparing local  health data with
 regional, State or national data so that any statistically signifi-
 cant increases can be detected

NECESSARY INFORMATION

  Public health officials must have specific information if they are
to conduct a proper  public  health evaluation of a hazardous site.
Reports for review need to be well organized  and concise  and they
should include summary tables of the  environmental data. When
appropriate, these summary tables should be compared  with  ap-
plicable Federal  and  State standards,  criteria  and  guidelines.
Volumes of raw data should be included in a separate appendix,
not in the report. The report should include:
•Characterization and  a brief history of the site, adjacent land
 uses and identified health problems
•Name and approximate quantity of the contaminants of concern
 on-site and off-site
•Topographic,  hydrologic  and  geologic  descriptions, as well as
 maps of the site (aerial photographs can be helpful)
•A description of the accessibility of the  contaminated zones on-
 site or off-site to human use and the land use (for example, resi-
 dential,  school,  recreational, commercial  fishing,  hunting,
 gardening,  playground, livestock grazing and crop production)
•A description of the land uses  and size of local populations that
 have been exposed or may have exposure risk
•A description of the quality, use  (that is, recreation, irrigation,
 livestock, drinking water,  food production, etc.) and proximity
 of local groundwater  and surface water resources; the presence
 of any  USEPA designations as sole-source aquifer should be
 noted
                                                                                   ENDANGERMENT ASSESSMENT
                                                          235

-------
•A description of the extent  of contamination and the rate and
 direction of contaminant migration in the various transport path-
 ways (air, soils, sediments, surface water and groundwater) on-
 site and off-site
•Description of fire and explosive hazards and potential adverse
 airborne releases
PRIORITIZATION OF HAZARDOUS
WASTE SITES

   CDC's S.P.A.C.E. for Health manual14 explains how to assign
and adjust the priorities of a hazardous waste site because of the
potential impact on human health. Procedures for using all infor-
mation gathered from the record, inspections and analyses to ob-
jectively assign site priorities in terms of public health are described
in the manual. This priority can be used to make effective decisions
about the funding of remedial actions, cleanup activities, further
environmental testing and similar matters. Priorities may later need
to  be adjusted because of the results of further environmental
testing, routine health monitoring, health studies or problems ap-
pearing during remedial actions.
   The priority classification of a site is based on characteristics
about the site, the contaminant of  concern, the  potential en-
vironmental pathways,  the potential for human exposure/absorp-
tion and  the health effects  in the exposed population.  The check
list'4 at  the end (Appendix 1) provides details in each  of these
categories.
                                                          CONCLUSION

                                                            A general outline of the items necessary for evaluating the public
                                                          health significance  of hazardous waste sites has been presented.
                                                          The PHS/CDC has been the nation's focal point for conducting
                                                          many of the public health evaluations of Superfund hazardous
                                                          waste sites. Because of the enormity of the hazardous  waste prob-
                                                          lem and  the potential public health threat,  identified hazardous
                                                          waste sites  need to  be evaluated in a more timely manner. These
                                                          potential  health problems justify the State's playing larger roles in
                                                          accomplishing these evaluations within budget constraints.
                                                            PHS/CDC intends to continue to be the  nation's focal point for
                                                          providing advice and support on the public health implications of
                                                          hazardous waste sites, for national  registries and  inventories re-
                                                          quired by CERCLA,  for conducting screening and health effect
                                                          studies  and for responding to  public  health emergencies due to
                                                          hazardous and  toxic substances. PHS/CDC also  intends to con-
                                                          tinue to act as an advocate for the resources that States need to par-
                                                          ticipate fully with CDC in this program. Only with the full coopera-
                                                          tion  of  each  State's public health  and  environmental/natural
                                                          resources agencies can  we make  real progress in removing the
                                                          public health  threats posed by the multitude of hazardous waste
                                                          sites and  spills.
                                                            PHS/CDC depends upon and encourages constructive feedback
                                                          from other Federal and State agencies to  improve it  evaluation
                                                          procedures, protocols, information  resources  and recommenda-
                                                          tions provided to each region and State. CDC encourages your con-
                                                          tinued comments, particularly on  its new S.P.A.C.E. for Health
                                                          manual.
                                                            Appendix 1
                                           Checklist for Use In Determining Priority of a Site


                    For each factor on the list, enter a check beneath the criterion  level (0, I, 2, 3, or unknown) that applies to the rite.

                                            Unknown             0             1             2_             )


                1.  HAZARDOUS SITE
                   DocunentatIon of pretence
                   of hazardoua site
                                   no date/
                                   DO conclusion
                                                           uncorroborated
               hi etorl cal
               records
               observation of
               wast*  releaee
                 laboratory
                 Co of 1 rma t loo
                   Toxlcity of 5 Boat
                   hazardous substance!
                   • t site
                   (App. t - Tables I. 2)
                                   no data/
                                   no conelualon
                                             high
               Ic.  Quantity of 5 eoet
                   hszardoua substances
                   at alt*
                   (App. B  Table 3)
                                  ao data/
                                  no conelusloa
                                             hi|h
                   hazardous substance*
                   at • 1 ta
                   (App. B - Tables 2. t)
                                            no data/
                                            no conelualon
                                                                                              tilth
                   Concentration of 5
                   • oat hazardous subatancee
                   (e.g..  In water and
                   soil on tIte)
                                  no data/
                                  no coneluslon
^ background
  levele
> background
 lev*la
greatly exceed
background
levels (or abow
standards)
above levela
with elgnl-
flcant hare.
poteotlal
               If.  Site BanagCEjent and
                   and contelraent
                   (App. B - Tablea 3, 6)
                                  ao data/
                                  no concluelon
total
control
adequate
control
Inadequate
control
                                                                                             uncontrolled
236
ENDANGERMENT ASSESSMENT

-------
Ig.  Potential for direct
     access to aite
                                   no data/
                                    occasional
                                    individual
                                   small popula-
                                   tion with
                                   Intermittent
                                   access
                                    large popu-
                                    lation with
                                    repeated di-
                                    rect access
2.   EXPOSURE POTENTIAL OF
    ENVIROIMENTAL PATHWAYS
2a.   Ground water
     (App.  B   Table  7)
 no data/
 no conclusion
                                                                                                           high
2b.  Surface water
     (App.  B - Tables  8. 9)
 no data/
 no conclusion
                                                                       low
                                                       high
2c.  Air
                                   no data/
                                   no conclusion
                   no suspected
                   releases
rare reported
apparent
effects



intermittent
infrequent
complaints



repeated re-
levels ex-
ceed stan-
dards; fre-
quent major
complaints
2d.   Deposition In (on)  soil
     off site
 no data/
 no conclusion
 absent or
  < background
 1eve1s
  > background
levels
 greatly
 exceed
 background
 levels
                                                                                                          above level
                                                                                                          with signi-
                                                                                                          ficant harm
                                                                                                          potential
2e.  Presence In food chain
                                  no data/
                                  no conclusion
                  absent
                  or  background
                  level
                  moderate In-
                  crease over
                  background
                  but  below FDA
                  standard
                at or near
                FDA standards
                   signifi-
                   cantly  above
                   FDA
                   standards
    POTENTIAL FOR HUMAN
    EXPOSURE/ABSORPTION
     Presence of  potentially
     exposed population  (I.e.
     people living, working
     or otherwise occupied
     In area near site or
     or relevant  pa thway)
no data/
no conclusion
no people
within 1
mile of
site or
relevant
pathway
people with! n
1 mile but
not In imme-
diate vicinity
(e.g., not
within 1-2
blocks) of
site or rele-
vant pathway
small number
of people
( < 100) in
immediate
vicinity of
site or
re levant
pa thway

                                   large number
                                   of people
                                   ( > 100)  in
                                   immediate
                                   vicinity of
                                   site or
                                   relevant
                                   pathway
3b.   Basis  of evidence for
     human  exposure/
     absorption
 no data/
 no conclusion
 unfounded
 allegations
historical
records
highly sugges-
tive data
from environ-
mental
monitoring
                                                                                                          results of
                                                                                                          biological
                                                                                                          sampling
                                                                                                          and/or
                                                                                                          presence of
                                                                                                          characteris-
                                                                                                          tic illness
                                                                                                          for relevant
                                                                                                          exposure
3c .   Levels  of  substances
     found  through
     biological sampling
                                   no data/
                                   no conclusion
                  substances
                  not detected
                  or  < back-
                  ground
                  levels
                  small* pro-
                  bably Inslgnl-
                  cant elevation
                  over back-
                  ground levels
                 significant
                 elevation,
                 clinical
                 effects
                 uncertain
                   exceed
                   levels with
                   potential
                   for  illness
                                                                                  ENDANGERMENT ASSESSMENT
                                                                                                  237

-------
                     HEALTH EFFECTS IN
                     EXP02D POPULATION
                      Allegatlona/reporti of
                      health effeed
                                      no data/
                                      no cone luslon
no a llega~
tione or
report*
vague, non-
•pecIt Ic ,
poorly charac-
t arlxed «lle-
gatlona
•peclflc and
w*l1-documented
report •, but
ef feei• of
doubt fuI rele-
venee (or cheal-
rale  und t r
contlderat ton
• olid re-
port! of
re levant
effaeca for
cheaUcale
under cen-
• Iderac Ion
                      Re»ult§ of clinical or
                      cpldealologlc atud lei
                      conducted
                                      no date/
                                      no cone lualon
good itud?
with negat 1
re eulli
                                                                                Halted  etudy
                                                                     or tnjlgnl"
                                                                     Meant re-
               •clentlfleallr
               Halted el«dy
               vich positive
               or Important
               finding!
                  •cleoilfl-
                  cally louftd
                  etudy with
                  poeitlve.
                  elgnlficant
                  flodlngt
                                      rre nt ly
                                                no data/
                      obte
                      shorr-tera) health effect
                                                                                                      high e«pee-
                                                                                                      tatlo* of
                                                                                                      C u r re nt
                                                                                                      effect
                      Expectation of a future
                      (often chronl c or long-
                      tera) health effect
                                                                                                      high expec-
                                                                                                      tation of
                                                                                                      future
                                                                                                      effect
                      S*verl ty or public
                      hea 1 th  lapac t  of
                      pr* »uo*d health
                      effee c •
                                      no data/
                                      no conclusion
                                                                negligible
               • Inletal
               health
               effecte, bu
               wide epread
               potentlally
               •evere  health
                 aewerc
                 health
                 eff.ct..
                 with vlde-
                 •pread
REFERENCES

  1. Sax, N.I.,  Dangerous  Properties of Industrial Materials, 6th Ed.,
    Van Nostrand Rheinhold Co., New York, NY,  1984.
 2. National Fire Protection Association,  National Fire Codes, 13, No.
    49, 1977.
 3. National Institutes of Health, U.S.  Public Health Service, National
    Toxicology Program Technical  Report Series,  and National  Cancer
    Institute Technical Report Series, 1976 to date.
 4. National Institute for  Occupational Safety and  Health (NIOSH),
    CDC, U.S.  Public  Health  Service,  RTECS, Registry of Toxic Ef-
    fects of Chemical Substances,  1981-82 Ed., Volumes 1 to 3, June
    1983.
 5. NIOSH, CDC, U.S.  Public  Health Service,  Occupational  Health
    Guidelines for Chemical Hazards, January  1981.
 6. National Library  of Medicine provides following services—CHEM-
    LINE, TOXLINE, RTECS, etc.
 7. USEPA, Office of Health  and  Environmental Assessment,  Health
    Assessment  Documents,  for  many chemical  substances, External
    Review Drafts.
 8. USEPA, Ambient Water Quality Criteria Documents, for substances
    listed in Nov. 28,  1980 Federal Register.
 9. National Academy of Sciences, National Research Council, Drinking
    Water and Health,  Volume 1 — 1977,  Volumes 2  and 3—1980, and
    Volume 4—1982.
                                                              10. USEPA, Aquatic Fate Process Data for Organic Priority Pollutants,
                                                                 Final Report, 440/4-81-014, Dec. 1982.
                                                              II. USEPA, "National Oil and Hazardous Substances Contingency Plan
                                                                 Title," 40 CFR Part 300, Federal Register, 47, No. 137. July 16, 1982.
                                                              12. USEPA. Water-Related Fate of 129 Priority Pollutants, Volume I:
                                                                 Introduction  and  Technical  Background,  Metals  and  Inorganics,
                                                                 Pesticides and PCBs; Volume II:  Halogenated Aliphatic Hydrocar-
                                                                 bons, Halogenated Ethers, Monocyltic, Aromatics, Phthalate Esters,
                                                                 Polycyclic Aromatic Hydrocarbons, Nitrosamines, and Miscellaneous
                                                                 Compounds, EPA'440/4-79-29 a, b, Dec. 1979.
                                                              13. U.S. Department of Transportation, Research and Special Programs
                                                                 Administration and  National Highway Traffic Safety Administra-
                                                                 tion. Emergency Action Guide for Selected  Hazardous Materials,
                                                                 1978.
                                                              14. Centers for Disease Control, U.S. Public Health Service, A System For
                                                                 Prevention, Assessment, and Control of Exposures and Health Ef-
                                                                 fects from Hazardous Sites (S.P.A.C.E. for Health], Jan. 1984.
                                                              15. USEPA, Water Quality Assessment: A Screening Procedure for Toxic
                                                                 and  Conventional Pollutants—Part  1  and Part  2, EPA 600/6-82-
                                                                 004,  a. b, Sept. 1982.
                                                              16. USEPA, (contracted  to Tetra Tech,  Inc., Lafayette, California),
                                                                 Water  Quality Assessment of Toxic and Conventional Pollutants in
                                                                 Lakes, Streams,  Estuaries and Ground Water, June  1984  Denver
                                                                 CO.
238
ENDANGERMENT ASSESSMENT

-------
CHILDREN'S EXPOSURE TO SMELTER-ASSOCIATED LEAD,
                                   MONTANA  AND IDAHO

                                            R. SCHILLING, D.V.M.
                                                 D. ROSS, Sc.D.
                                                D.  SOKAL,  M.D.
                                                  R. ING, M.D.
                                       Center  for Environmental Health
                                          Centers  for Disease Control
                                                Atlanta, Georgia
                                             C. BROKOPP, D.P.H.
                                  Idaho Department of Health and Welfare
                                                   Boise, Idaho
                                                A.D. MAUGHAN
                        Montana Department of Health and Environmental Sciences
                                                Helena, Montana
INTRODUCTION
  Lead smelting has been a major industry in East Helena, Mon-
tana, and Shoshone County, Idaho, since the turn of the century.
In both places, emissions from the smelters have contaminated
nearby residential communities. In 1974, the average air lead level
within 1 mile of the smelter in East Helena was four micrograms
per cubic meter,1 and the level within  1 mile of  the smelter in
Shoshone County was 17 micrograms per cubic meter.2 For com-
parison, the present national ambient air quality standard is 1.5
micrograms per cubic  meter.3
  Children living in East Helena and in Shoshone County in 1974
and  1975 had elevated blood lead levels. Although samples were
collected during the winter, when exposure to soil lead is minimal,
10% of the children in  East Helena had whole-blood lead levels
greater than 40 micrograms per deciliter Otg/dl), meeting the ac-
cepted  definition at that time for elevated blood lead/ Summer
sampling in Shoshone  County in 1974 indicated 42% of the
children had blood lead levels above 40 /*g/dl.!
  The  findings in Idaho resulted in remedial actions aimed at
reducing smelter emissions and limiting access to contaminated
soils. Included were:  (1) closing the smelter temporarily; (2) im-
plementing  smelter emission controls;  (3)  improving  industrial
hygiene practices at the smelter to reduce the amount of lead dust
brought home by workers; (4) purchasing and destroying homes
near the smelter after relocating residents to less-contaminated
neighborhoods; (5) covering contaminated yards with imported soil
and seeding these yards with grass; (6) providing water subsidies to
promote the growth of the grass; (7) covering driveways and play
areas with new sand  and gravel; and (8) providing community
education programs on how to reduce children's exposure to con-
taminated  soil and household dust. In 1981, air lead levels in
Shoshone  County,  Idaho, decreased markedly when the lead
smelter closed.
  In East Helena, Montana, remedial actions were taken in 1982 to
reduce  exposures to  lead-contaminated dust,  including  (1)  im-
plementing a street-cleaning program; (2) planting trees around the
smelter; and (3) building concrete bins for storing lead ore.
  In 1983, to determine whether a public health problem persisted
in East  Helena and Shoshone County, questionnaire  data and
biologic and environmental samples for evaluating the children liv-
ing in those smelter communities was collected. The study purposes
were twofold. The first purpose was to identify children who had
excessive blood lead levels with or without evidence of lead toxicity.
Since 1974,  new epidemiologic, clinical and experimental evidence
has indicated that lead is toxic at levels  previously thought to be
non-toxic.6"13 Currently, the Centers for Disease Control is revising
its criteria for lead toxicity. The proposed levels are blood levels
greater than or equal to 25 /ig/dl and erythrocyte protoporphyrin
(EP) greater than or equal to 35 /tg/dl. The direct toxic effect of
lead on heme biosynthesis is seen in increased levels of EP in whole
blood.  An elevated EP level thus can provide early and reliable
evidence of biochemical toxicity due to lead.
  The second  purpose of the 1983 epidemiologic studies in Mon-
tana and Idaho was to determine the relative contributions to
children's blood lead in these communities from lead-contaminated
air, soil and dust and from host factor characteristics such as play
habits and the occupations  and hobbies of family members.

METHOD
  Using geographic patterns of environmental lead contamination,
three study areas in each community were defined. Area 1 included
homes  within  a 1-mile radius from the smelter.  Area 2 included
homes from 1  to slightly more than 2 miles away. Area 3, the com-
parison area, consisted of a nearby neighborhood selected for hav-
ing age, race and socio-economic characteristics similar to those of
Areas 1 and 2. In both Montana and Idaho, Area 3 was about 5
miles west of the smelter, a direction which is usually upwind.
  A door-to-door census was used in each study area to identify
which families had lived in the community for 3  months or more
and, in Montana, had children 1-5 years old and in Idaho 1-9 years
old. Older children were included in Idaho so that the results could
be compared with results of previous studies.  Survey teams visited
the eligible households with questionnaires, collected environmen-
tal  samples,  and, in  Montana, drew  blood samples  from the
children. In Idaho, blood samples were collected at a local clinic.
  The questionnaire design used in Montana and Idaho permitted a
parent to select the best of several preceded responses to describe
the habits of  individual children and  the characteristics of the
household. Questions on each child's habits addressed (1) oral ac-
tivity (e.g., whether the child ate snow, used a pacifier, sucked a
thumb, chewed fingernails, carried a favorite blanket or toy around
during the day and put this in his or her mouth, placed his or her
mouth on furniture, placed paint chips or other objects in his or her
mouth, or swallowed foreign objects);  (2) play environment (e.g.,
did the child play with other children  or adults,  stay at home to
play, play indoors or outside, take food outdoors during playtime,
spend time on the floor, play on grass, dirt  or concrete surfaces
outside?); (3)  routine hygiene practices ( e.g.,  were the child's
hands usually  washed before eating, before going to sleep, and
after playing with dirt or sand?); and (4) dietary  supplementation
with minerals and vitamins. Questions on the characteristics of the
household addressed  socio-economic  status, length of time in
                                                                             ENDANGERMENT ASSESSMENT
                                                      239

-------
residence,  lead-related  occupations  and  hobbies  of  family
members, use of fruits and vegatables grown in the neighborhood,
food storing and serving practices that might increase the lead con-
tent in the family diet; year of house construction; and presence of
a smoker in the household.
  To identify the sources of lead in  the children's usual surroun-
dings, the survey teams collected composite samples of soil from
front and back yards  of each house. Samples of household dust
were  also collected,  using  both floor wipes  and samples from
vacuum   cleaner  bags.   When  available,  samples  of  garden
vegetables were collected  for heavy metal analysis. To determine
the amount of  lead present  in painted surfaces inside and  outside
each  home, hand-held X-ray fluorescence (XRF) analyzers were
used. Data on ambient air lead were collected throughout the study
at air monitoring stations in each of the three study areas.
  Blood samples  were  analyzed  by   ESA  Laboratories,
Massachusetts, for blood lead, EP and hemoglobin. EP concentra-
tion was determined  by  the extraction  method.  Environmental
samples  (soil,  dust and  vegetables)  were analyzed by the Silver
Valley Laboratory, Idaho, and the Montana Department of Health
and Environmental Sciences.

RESULTS AND DISCUSSION

Blood-Lead and EP

  The numbers of eligible children, i.e., those who met the age and
residence criteria, were 437 in Montana and 400 in Idaho. Tables 1
and 2 give the number  of eligible children and the number of
children tested  in each study area in Montana and Idaho. Participa-
tion rates in Montana were highest in Area  1  (95%), the Area
closest to the smelter, and lowest in Area 3 ( 77%), the comparison
area. In  Idaho, Areas 1 and 3 both had participation rates of 94%.
  Tables 3 and 4 give the results of children's blood lead and EP
levels in Montana and Idaho. The log-transformed mean blood
lead  data  in  both   areas  were consistently  and  statistically
significantly higher for children living near the smelter (p c 0.001).
In  addition,  children  in  Idaho  Area 1 who were born after the
smelter closed  in 1981 had a mean blood lead level of 22 pg/dl.
  Six percent of the children who lives in Area 1 in Montana had
blood lead values that showed excessive lead absorption (i.e.,   25
/tg/dl). No child who lived outside Area 1 in Montana had a blood
lead value of 25  /ig/dl or greater. In comparison, relatively high
                            Table 1
          Numbers of Eligible and Tested Children, Montana
                            Table 3
Blood Lead and Erythrocyle Proloporphyrln (EP) Levels, Montana (MS/dl)
Area
1
2
3
Total
Dili, from
Smeller (miles)
<1
1-2.25
>5

No. of Eligible
Children*
104
254
79
437
No. of Children
Toted
98 (95%)
238 (94%)
60 (76%)
396(91%)
*I-5 years of age
                           Table 2
           Numbers of Eligible and Tested Children, Idaho

Area
1
2
3
Total
Distance from
Smeller (mile*)
<. 1
1-2.25
>5

No. of Eligible
Children*
46
223
131
400
No. of Children
Tested
43 (94%)
199 (90%)
122 (94%)
364 (91%)
Blood Lead
Area
1
2
3
Mean
13
9
6
«* 25 N*
6 6
0 0
0 0
Erylhrocyte Proloporphyrin
Mean % 35
22 1
21 4
20 3
N"
1
9
2
No. of
Children
«i(h Lead
Toxkityt
1
0
0
•Number of children with blood lead
••Number of children wilh EP  »l/dl
tDlood lead  23 c»/dl and EP  33 *f/dl

                           Table 4
Blood U»d and Erythrocyle Protoporphyrin  EP )! .i/dl
(Blood lead 2] n/t\ «nd tP )J ft/a

percentages of children in Idaho Area 1 (35%) and Area 2 (15%)
had blood lead values greater than 25 pg/dl.
  The  log-transformed EP levels in Idaho were consistently and
significantly higher for children living near the smelter (p   0.05).
No significant differences were found in mean EP levels among the
three study areas  in Montana.

                           TableS
         Mean Air Lead Levels Oifc/nP), Montana and Idaho
                      August-October 1983
Area
1
2
3

Mean Lead
Area
1
2
3
Montana
3.70
0.98
0.23
Table 6
Levels (ppm*) la Soil and
Soil
6,059
3,432
677
Idaho
0.28
0.12
0.10

House Dust, Idaho
House Dust
4,136
4,875
1.361
*l-9 years of age


240       ENDANGERMENT ASSESSMENT
• parts per million


  Children with  iron deficiency  anemia  may  have elevated EP
values in the absence of lead exposure. To determine whether the
similarity of EP levels among the three  study  areas  in Montana
could be due to a confounding effect of  iron deficiency anemia,
blood  hemoglobin data were analyzed. Out of 3% children who
had blood samples analyzed for EP, only  three had low (i.e., less
than 11 grams per deciliter) hemoglobin concentrations.
  One child in Montana and 30 children in Idaho had lead toxicity,
as defined by having both a blood lead level of at least 25 pg/dl
and an EP value of at least 35 ug/dl. No child had high blood lead
levels (i.e.,  > 55>ig/dl) that required chelation therapy.

-------
                            Table 7
       Mean Lead Levels (ppm*) in Garden Vegetables, Idaho
                             Table 8
   X-Ray Fluorescence (XRF) Analysis of Lead in Household Paint, Idaho
Area
1


2


3


Total
Vegetables
Carrots
Beets
Lettuce
Carrots
Beets
Lettuce
Carrots
Beets
Lettuce

Mean Lead
Level
18
0
48
61
47
65
25
16
32

No. of Samples
3
0
1
23
14
25
32
9
20
127
 *parts per million
Air Samples

  Table 5 gives the mean air lead levels during the study period in
Montana and Idaho. The mean air lead levels in Areas I and II in
Montana exceeded those in Idaho by a factor of 10; this difference
reflects the absence of active smelting in Idaho.
Soil and Dust
  Laboratory analyses of environmental samples (soil, dust, and
vegetables) from Montana have yet to be completed. Table 6 gives
the mean lead levels found in soil and house dust in Idaho. Within
each study area in Idaho, soil lead levels collected from  adjacent
yards varied greatly. No concentric pattern  or other consistent dis-
tribution of lead in the soil of the three study areas were noted.
The absence  of such a pattern may be due  to the remedial actions
taken in 1975 when selected residential yards were covered with im-
ported soil and selected playgrounds were  covered with new sand
and gravel. Those actions may also account  for the fact that in 1983
the highest values in samples of yard soil (41,200 ppm) and of gar-
den soil (5,160 ppm) were found in Area II rather than in Area I.
  Samples of house dust taken from vacuum cleaner bags showed
higher lead levels in Idaho Area I (4,136 ppm) and Area  II (4,875
ppm) than in Area III (1,361 ppm) by about  threefold.

Garden Vegetables
  Table 7 gives the mean lead levels found in garden vegetables
in Idaho.  Most of the samples were collected in Areas II and HI.
Although lead levels in carrots, beets, and lettuce were about half
those found  in 1974 and 1975,  the  average levels in Area II re-
mained two to three times higher than the levels in Area III.

Lead in Paint
  Table 8 gives the X-ray fluorescence (XRF) data on lead concen-
trations in household paint in Idaho. These data show a low prev-
alence of leaded paint in the homes of study participants. No corre-
lation was found between children's high  blood lead  values and
residence in homes with high lead paint values.

Questionnaire
  The questionnaire  data  are  now being  analyzed. A  major
hypothesis to be tested is the relationship between children's oral
activities and blood lead levels.

SUMMARY

  The  1983 epidemiologic study of children's exposure to lead in
East Helena, Montana, showed  that children who lived  near the
smelter had higher blood lead levels  than those who lived farther


Lead Category
Negative
Low
Moderate
High

XRF Reading
(mg/cm1)
<0.7
0.7-2.9
3.0-5.9
> 6.0
Number
of Surfaces
Tested
1,133(78%)
186(13%)
75 ( 5%)
66 ( 4%)
                                                                                                               1,460
away. After completing the laboratory analyses of environmental
samples, the relative contributions to the blood lead levels from the
lead levels in air, dust, and soil and from the children's play habits
and family characteristics will be determined.
  The 1983 epidemiologic study of children's exposure  to lead in
Shoshone County showed that children who lived in areas with high
levels of lead in the soil were more  likely to have elevated blood
lead levels than children living in areas  with  cleaner soils. Specif-
ically, within 1 mile of the smelter, where the mean soil lead  level
was 6,059 parts per million, the blood lead level of 35% of the chil-
dren was 25'jug/dl or higher. These blood lead levels were consid-
ered to be excessively high, and they may result in adverse health
effects.
  Since a child who eats a half-cup serving of vegetables contain-
ing 15 ppm of lead would be ingesting an amount of lead exceed-
ing the  recommended maximum daily intake, it is recommended
that children not eat leafy or root vegetables grown in any of the
three study areas in Shoshone County, Idaho.
  Of special interest in Idaho was the finding that 1- and 2-year-
old children within 1  mile of the smelter had a mean blood  lead
level of 22 pg/dl. Since these 11 children were born after the smelter
closed in 1981, they have not been exposed to high air lead levels.
Unless they had been exposed to some unusual source of lead, these
children would be expected to have a mean blood  lead level of
about 6 ug/dl. Since air lead levels in Idaho in 1983 were low and
since very little leaded paint was found  in the homes studied, the
main source of lead exposure probably has been the contamination
in the dust and soil. This  hypothesis will be tested in multivariate
regression analyses.
  The environmental sources of lead in 1983 in Shoshone County,
Idaho, were more diffuse than those in 1974, and the  children's
blood lead levels were lower. It is suspected that these characteris-
tics of the data may substantially limit the predictive  power of the
final multivariate model. Thus, the final model for explaining  chil-
dren's blood lead levels in Shoshone County may leave most of the
variance in these levels unexplained.

REFERENCES

 1. Montana  Department  of Health and Environmental  Sciences.  Air
   Quality Bureau Records. Helena, Montana. 1974.
 2. Shoshone Lead Health Project: Work Summary, Idaho Department of
   Health and Welfare, Division of Health, Boise, Idaho, 1976.
 3. Environmental Protection Agency: Air Quality  Criteria  for Lead,
   Washington, DC: U.S. Environmental Protection Agency, Office of
   Research and Development, Publication EPA—600/8-77-017,1977.
 4. "Medical aspects of childhood poisoning, "Pediatrics, 45:464,1971.
 5. Landrigan, P.J., Baker, E.L., Feldman, R.G.,  Cox, D.H., Eden,
   K.V., Orenstein,  W.A., Mather, J.A., Yankel,  A.J. and Von Lindern,
   I.H., "Increased lead absorption with anemia  and slowed  nerve con-
   duction in children near  a lead smelter", The Journal of Pediatrics
   89(6): 904, December, 1976.
 6. Piometti, S., Seaman, C., Zullow, D., Curran, A. and Davidow, B.,
   "Threshold  for lead damage to heme synthesis in urban  children"',
   Proceedings of the National Academy  of Sciences, USA 79'333s'
   1982.
                                                                                   ENDANGERMENT ASSESSMENT
                                                          241

-------
7. Rosen,  J.F.,  Chesney, R.W., Hamstra,  A., DeLuca, H.F.  and
   Mahaffey, K.R., "Reduction in 1,25-dihydroxyvitamin D in children
   with increased lead absorption", New England Journal of Medicine,
   302:1128,1980.

 8. Angle, C.R. and Mclntire, M.S., "Low level lead and inhibition  of
   erythrocyte pyrimidine nucleotidase", Environmental Research,  17:
   296,1978.

 9. Paglia, D.E.,  Valentine, W.N. and Fink, K., "Further observations
   on erythrocyte pyrimidine nucleotidase deficiency  and intracellular
   accumulation of pyrimidine nucleotides", Journal of Clinical Investi-
   gation, 60:1362, 1977.

10. Needleman, H.L., Gunnoe,  C., Leviton, A., Reed, R., Peresie, H.,
   Maher, C. and Barrett, P.,  "Deficits in psychologic and classroom
                                                                 performance of children with elevated dentine lead levels", New Eng-
                                                                 land Journal of Medicine, 300:689, 1979.
                                                             11. Yule,  W., Lansdown, R., Millar, I.E. and Urbanowicz, M.A., "The
                                                                 relationship  between blood lead concentrations,  intelligence and
                                                                 attainment in a school  population: a pilot study". Developmental.
                                                                 Medicine and Child Neurology. 23:567, 1981.
                                                             12. Benignus, G.A., Otto, D.A., Muller. K.E. and Seiple,  K.J., "Effects
                                                                 of age and body lead burden on CNS function in young children. II.
                                                                 EEG spectra", Electroencephalography and ClinicalNeurophysiology,
                                                                 52:240,1981.
                                                             13. Seppalainen, A.M. and Hernberg, S.,  "A follow-up study of nerve
                                                                 conduction velocities in lead exposed workers", Institute  of Occupa-
                                                                 tional  Health, Helsinki, Finland,  Neurobehavioral Toxicology and
                                                                 Teratology, 4:721, 1982.
242
ENDANGERMENT ASSESSMENT

-------
       THE USE OF  SERUM  REFERENCE MATERIALS  AND
     STATISTICAL  METHODS IN THE CLASSIFICATION  OF
           HUMAN  EXPOSURE TO  PCBS  AT  WASTE  SITES
                                            VIRLYN W. BURSE
                                         JOHN M. KARON, Ph.D.
                                       DOUGLAS M. FAST, Ph.D.
                                         JOHN A. LIDDLE, Ph.D.
                                     Center for Environmental Health
                                         Centers for Disease Control
                                              Atlanta,  Georgia
INTRODUCTION

  The improper dumping of waste from various industrial pro-
cesses into landfills, waterways and rock quarries has created toxic
waste sites throughout the United States. In 1980, Congress passed
CERCLA to provide money for  finding, cleaning  up and con-
trolling these toxic waste sites. State agencies and the USEPA have
worked together to  find these sites and rank them on the basis of
the type and quantity of waste on the site and the potential risk
posed to public health.
  The Centers for  Disease Control (CDC) has been involved in
assessing the health  of persons living near specific waste sites. The
model' shown in Figure 1 illustrates the potential ways that toxic
chemicals  found in  waste sites can get into environmental path-
ways—which are simply soil, air, water and the food chain. Human
exposure can result  from ingestion, inhalation and/or dermal ab-
sorption.
  Human exposure has to be proven. It cannot be assumed. It must
be shown that humans bodies contain a particular chemical or its
metabolite that is found at a specific dump site. This is often diffi-
cult for three reasons:
•Many of the chemicals  found at the dump site have short half-
 lives in mammals.
•People may have had only intermittent exposure to the dump site,
 resulting in exposure levels below the detection limits of available
 analytical methods.
•Most people have  low "normal" background levels  of certain
 persistent chemicals from exposure through the food chain.
                     Watte site
                     containing
                     toxic chem
                                Contaminated
                                environmental
pathway*
(soil, water.
lood. etc I
Human Exposure
(through skin
contact, mges
non, inhalation)
                            Certain chemicals found in waste sites have these characteristics:
                          (1) they have long half-lives; (2) they are stored in human bodies;
                          (3) they can be identified analytically; and (4) continuing low-level
                          exposure to them could result in adverse long-term health effects.
                          Polychlorinated biphenyls (PCBs) have these characteristics.  In
                          addition,  their presence at waste sites is so widespread that the
                          Mitre Corporation used PCBs as part of the basic criteria for rank-
                          ing hazardous waste sites.2
                            The amount of PCBs in a person can be estimated by analyzing
                          serum and fatty tissue. Since the average amount of PCBs in the
                          general American population has been estimated, comparisons can
                          be made between the body burdens of persons exposed to specific
                          waste sites and the general population.
                            CDC personnel experienced in evaluating health problems con-
                          nected with waste sites have designed a pilot study for use in coop-
                          eration with local health departments and the USEPA, to deter-
                          mine the amount (body burden) of PCBs in persons living close to
                          the 20 highest-ranking waste sites. CDC personnel also have devel-
                          oped—and described in a paper not yet published—a multistage
                          plan that would be cost effective because only  those  individuals
                          with the  most  significant PCB-exposure would receive further
                          study. The suggested four stages of this study are:
                          •Site ranking. Review environmental and population data on all
                           toxic waste sites with high PCB levels in pathways to which peo-
                           ple could be exposed, and select for the  pilot studies those sites
                           with the greatest potential for human exposure.
                                                                Acute body burdens
                                                                (chemical! detected
                                                                only during or
                                                                shortly alter   	
                                                                exposure)
                                                                Cumulative body
                                                                burdens (chemicals
                                                                detectable for
                                                                long periods after
                                                                cessation of exposure)
                                                 Acute (short
                                                 term) health
                                                 effects —
                                                 signs and symptom
                                                 appearing shortly
                                                 after exposure
                                                  Subacute. chronic
                                                  or latent (long
                                                  term) health effects-
                                                  including repro
                                                  ductive and develop
                                                  menial abnormalities
                                                  other chronic
                                                  diseases and cancer
                                                     Figure 1
   Model of Potential Human Health Effects Related to Chemical Exposures from Toxic Waste Sites from E.R. Welty et a/., Unpublished Data, CDC.
                                                                         ENDANGERMENT ASSESSMENT
                                                                              243

-------
 •Pilot exposure studies. By using questionnaires and environmental
  data, select the most heavily exposed persons chosen in Stage  1
  and measure their serum PCS levels. In instances of exposure of
  lactating mothers, analyze breast milk for PCBs.
 •Community surveys. At those sites where Stage 2 studies showed
  elevated serum PCB levels, conduct cross-sectional,  community-
  wide studies to determine health effects.
 •Cohort studies. On the basis of data from Stages 2 and 3, register
  persons with markedly elevated PCB levels for long-term follow-
  up to evaluate any chronic health effects.
   These studies are intended to: (1) evaluate the health risk of per-
 sons living near dump sites; (2) indicate the proper health meas-
 ures to be taken; and (3) provide a model for assessing other  haz-
 ardous substances. CDC also  recognizes a need for a qualitative
 way to find the source of the PCB exposure. One method is to use
 specifically prepared serum reference materials. This study could be
 done simultaneously with the other investigations.
   PCBs are usually found in a metabolized or degraded state, and
 the nature of the original PCB exposure is difficult to determine.
 This is why basic reference material is so important.  In the work
 described here, specific PCBs are given to test animals, and the re-
 sulting blood samples are used as reference material. Suspected
 human PCB exposure can be checked by multivariate analysis to
 compare the relationship between the human sera samples and the
 reference materials. Analysts have attempted to quantitate PCBs
 using Aroclor® standards that produce patterns similar to or com-
 patible with the matrix pattern.
   For congener-specific identification of PCBs, some analysts, as
 well as members of the CDC analytical group, have used  capillary
 gas chromatography (CGC) with electron capture detection (ECD)
 or CGC with gas  chromatography mass spectrometry (GC-MS).
 The main problem with  this approach is the lack of standards for
 comparison. The PCB data in this report were obtained  by using
 packed-column gas chromatography (PCGC),  ECD  and Webb-
 McCalP mean-weight percent values.

 PREPARATION OF REFERENCE MATERIAL

   Before the serum reference material was prepared, the following
 decisions were made:  (1) which Aroclor® (AR) series were to be
 dosed; (2) which animal species was to be dosed; and (3) what dos-
 age level was to be used. The Aroclors®  were chosen on the basis
 of domestic sales since 1971 and how they were used—that is,  in a
 closed system, a normally closed system or an open-ended applica-
 tion.4 On the basis of this information, four Aroclors*  were fed
 singly. They were  AR 1016, AR  1242,  AR  1254  and AR 1260.
 Those fed in  combination were AR  1254/1016,  AR 1254/1260,
 AR 1254/1242 and AR 1242/1260.
   Adult female goats  were selected for the tests because of min-
 imum cost. The dosage level ( 100 mg/kg each Aroclor® ) was used
 because an earlier study had shown that this dosage provided an
 approximate 100 /tg/1 serum level in 30 days. Each goat was first
 tested for Aroclors®  and  chlorinated hydrocarbons (as DDT,
 DDE and others). The serum showed no significant  amounts of
 any of these chemicals. In this way, each goat served as its own
 control.  The goats then were force-fed gelatin capsules contain-
 ing the neat Aroclors®  . Each goat otherwise was fed a normal diet
 and kept in an individual stall.
   At the time of dosing, it was discovered that all the goats were
 pregnant except the one fed AR 1254. About 30 days after the kids
 were born, blood samples were taken from the mothers. The pro-
 cedures used in  the preparation of these samples are outlined in
 Table 1.
                           Table 1
             Preparation of Reference Serum Materials
 1. Female goats—average weight—41.9 kg
 2. Dosed orally (capsules) with a single  Aroclor (100 mg/kg) or a com-
   bination of two Aroclors each at 100 mg/kg
 3. 30 days later each goat was exsanguinated
   A. Serum obtained
   B. Sterile filtered (millipore)
   C. Dispensed (2-3ml) into screw cap vials
   D. Stored at -40°C
GAS CHROMATOGRAPHIC (GO PATTERNS

  The GC-ECD profiles  of the Aroclor*  standards used in the
feedings are shown in Figure 2. The GC-ECD profiles obtained
from sera after dosage ( >30 days) are shown in Figure 3 (single)
and Figure 4 (combination). The analytical method used to obtain
the GC traces has been reported,' except that the GC column temp-
erature was increased from 190°C to 205 °C.
  For single Aroclors®  (in vivo), peaks are identified with relative
retention times (RRT) to  DDE (DDE x 100).  In the combination
Aroclors®  (in vivo), peaks are identified as possibly being derived
from a particular Aroclor* or Aroclors* on  the basis  of the GC
traces obtained when only one Aroclor8 was fed.
  The concentration of certain Aroclor *  peaks differed notably
between the dose with one Aroclor*  and  those  with combina-
tions (Figs. 3 and  4). These differences appear primarily for the
early eluters in AR 1016  and AR 1242.  It is  not known whether
these apparent differences in concentration are due to the pharma-
codynamics of feeding two dissimilar Aroclors1 in combination
or to the nature of the ECD response to these types of compounds.*
This problem  will be investigated later.
                    SINGLE AflOCLORS DOSED
     1016     1242      1254
                                      1260
                                                                 1016      1242
                                                                                       1254
                                                                                                          1260
                          Figure 2
  Gas Chromatographic Profile (Electron Capture) of Aroclors"" Fed

244      ENDANGERHENT ASSESSMENT
                          Figure 3
     Gas Chromatographic Profile of Single Aroclors* > 30 Days
                        After Dosage

-------
        1242/1260
                                       1254/1260
                          Figure 4
   Gas Chromatographic Profile of Combined Aroclors®  i 30 Days
                        After Dosage

 SELECTED INCIDENCES OF HUMAN EXPOSURE

   Human sera from three cases of alleged PCB exposure provided
 samples  for comparison with the reference materials from the
 goats. The cases of alleged human exposure are outlined in Table
 2.
   The human serum tests can be compared visually with the goat
 reference serum tests in Figures 5-7. Note the similarities between
 the RRT of the human cases and the reference sera. Each case (ex-
 cept C) is shown with the in vivo Aroclor®  profile that  "best"
 relates to the Aroclor®   used to quantitate the case.  Case C, al-
 though quantitated as  AR 1260, contained early  eluting peaks
 (pre-DDE) indicating possible exposure to AR 1016 or AR 1242.

                           Table 2
             Selected Cases of Alleged Exposure to PCBs
Case
A
B
C
Background
A farmer who used Coumar®
as a sealant for his silo.
A composite sample from a Kansas family
that had consumed contaminated beef.'
A composite sample from residents
of New Bedford, Massachusetts.'
Quantitated as
Aroclor (ppb)
1254(172)
1260 (307)
1260(31)
DATA ANALYSIS OF GOAT AND HUMAN SERA
  Serum from each dosed goat was analyzed four times. Serum
from each human case was analyzed  once. Gas chromatograms
were characterized by relative retention times with respect to (DDE
x 100), resulting in the identification of 59 peaks.
  The responses for all peaks were computed as  area-%/mg of
serum injected. Peaks with responses of less than 0.25% (an arbi-
trary cutoff) were deleted, and peaks with relative  retention times
within ±5%  were combined. This resulted in 28 distinct peaks
present in at least one sample; that is, in a sample from one of the
three human cases or in a sample from  one of the eight goats used
in the 32 analyses.
  Although other statistical techniques—for example, SIMCA'—
are available for handling data of this kind, only two will be used.
The first, principal components10- u, will be used to  analyze the
area-percent data. This  is a technique for finding a new coordinate
system so that most of the variability in  the original 28-dimensional
space is (in this case) in the 2-dimensional plane defined by the first
two principal components.
                                                                              CASE A
                                                                                   3*7  410 "7    617
                                                                                    I 375    I
                                                                                    ill         .1
                            r
1254
                        Figure 5
 Gas Chromatographic Profile of Case A and In Vivo AR 1254 i 30
                    Days After Dosage
   144
 100  , 167  261
  tit    »   CASE B
44
1
168
1
t
191
I2"
1


307
J
337
t 477
l\ 410 t
vU t\
                                                                         MOT DETECTED BY MTEOHATOR
                                                                                                          1260
                                                                                           Figure 6
                                                                    Gas Chromatographic Profile of Case B and In Vivo AR 1260 >30
                                                                                       Days After Dosage
                                                                        I   t
                              1016
                       Figure 7
Gas Chromatographic Profile of Case C and In Vivo AR 1016 and
             AR 1242 > 30 Days After Dosage
                                                                                 ENDANGERMENT ASSESSMENT       245

-------
  The  second technique uses only the presence or absence of a
peak at each of the 28 relative retention times as the basis of analy-
sis. For each pair of gas chromatograms, a Jaccard coefficient11
was computed to measure the similarity between the gas chromato-
grams. This coefficient is the number of relative retention times
for which both gas chromatograms showed a peak, divided by the
number of relative retention times for which at least one of the two
gas chromatograms has a peak. Thus, this measure of  similarity
is between zero (no peaks in common) and 1 (identical peaks).
  The similarity between the different  analyses  done on serum
from a single goat, between the analyses of sera from two different
goats or between serum from a goat and serum from  a human
case was computed as the average of the similarities between the
corresponding pairs of gas chromatograms. In particular, the sim-
ilarity between the four gas chromatograms done on serum from a
single  goat  was  computed as the average of six  coefficients ob-
tained from all  possible pairs of  gas chromatograms. Thus, this
quantity  describes the stability of the  measurement process  on
different aliquots from a single animal.
   Similarly, the agreement  between the  gas chromatograms done
on sera from two different goats was obtained  by averaging 16
coefficients; and the agreement between the gas chromatograms
done on  serum  from a goat and  serum from a human case was
obtained by averaging four coefficients. This  measure  of agree-
ment is analogous to a correlation coefficient.
   The results of the principal components analysis are  shown in
Figure 8. This scatter plot  shows  the coordinates  for each gas
chromatogram with respect  to the two coordinate axes  (in 28-di-
 mensional space) defined by the  first two principal components.
 The gas  chromatograms corresponding  to the goats fed single
 Aroclors®  are separated and distinguishable on  the scatter plot.
 Gas chromatograms of sera from goats fed  combination Aro-
 clors® tend to be more similar to  the pattern corresponding to the
 Aroclor®  with the higher chlorine content. This  is not  surprising
 when  the stability of the higher chlorinated Aroclor®  is consid-
 ered.
   The human sera evaluated by this technique are not as clearly
 delineated. Serum from Case A,  a suspected 1254 exposure, does
 appear similar to 1254." Composite sera from Case B, a suspected
  1260 exposure, do not appear as  similar to 1260, however, as the
 visual comparison of the gas chromatographic analysis (see Figure
 6) would indicate. Composite sera from Case C,  although quanti-
 tated  as AR 1260, produced early eluting peaks (pre-DDE) which
 are considered indicative of exposure to AR 1016 and/or AR 1242.
 The principal components analysis indicates that Case C sera are
 more similar to AR  1016 and AR 1242 than to the other Aro-
 clors® .
                                                                                            Table 3
                     Similarity Measures* of Gas Chromatograms for Dosed Goals and Cases
                               1242/1254
                                         CASE A
                    CASE C
1254/1260

    ©
                               CASE B       1242/1
                      PRINCIPAL COMPONENT 2
Dosed
Goats
1260
1242/1260
1254/1260
1254
1242/1754
1016/1254
1242
1016
Cat A
Case B
Ca«C

1260
1 00
68
67
41
33
32
16
OS
37
62
42
1242/
1260
68
96
68
50
60
52
45
28
45
46
70
1254/
1260
67
68
96
72
60
59
35
IS
51
40
61

1254
.41
50
72
1 00
80
81
48
22
53
31
59
1242/
1254
33
60
60
80
96
87
66
34
51
29
74
1016/
1254
32
52
59
81
87
92
60
28
51
29
66

1242
.16
45
35
48
66
60
90
30
34
12
49

1016
08
28
15
22
34
28
30
61
18
12
31
                            Figure 8
       Principal Components Analysis for Goats and Selected Cases
 •Similarity meuura arc average* of J«cc«rd coeffkierm. u defined in Ihc leu.

  The Jaccard measures of similarity are  given in Table 3. They
show the same general pattern as the principal components analy-
sis, with some interesting differences. The agreement for the differ-
ent gas chromatograms from a single goat is excellent (at least 90%)
except  for the goat fed 1016; one of the four gas chromatograms
from the latter showed  many abnormal peaks.  Deletion  of that
aberrant gas chromatogram raised the average similarity among the
gas chromatograms of the goat fed 1016 to 80^o but had little effect
on the other similarity estimates for human cases or for the other
goats. Gas chromatograms from pairs of goats fed the same Aro-
clor®  (one alone, one in  combination with another Aroclor* )
generally show high similarity (at least 60Vo), except for the 1016
versus  1242/1260 pair (similarity 45%). This anomaly was also
detected by comparing the gas  chromatograms visually (Figs. 3
and 4).
  In general, the similarity patterns found among  the gas chrom-
atograms from the goats are similar to the patterns suggested by
the principal  components  analysis.  The  principal components
analysis suggests, however, that AR 1016 and AR 1242 are similar,
whereas the similarity measures  show an  agreement of only 30ro
between the gas chromatograms for these Aroclors* .
  The  similarity measures  suggest that the gas chromatograms
from Case A  are most similar to the pattern found in  1254, as did
the principal components analysis. This  simple measure of similar-
ity identifies Case B as a suspected exposure to 1260, in agreement
with the alleged exposure, but the principal components  analysis
does not make such  a clear identification. Finally, the similarity
analysis suggests that the gas chromatograms from Case  C most
closely resemble the patterns from AR  1242/AR 1254, AR 1242'
AR 1260 and AR 1016/AR 1254 in contrast to the principal com-
ponents analysis, which suggests that the gas chromatograms from
this case are most like AR 1242 alone. The composite sample from
Case C comes from residents of New Bedford, Massachusetts, who
may have consumed contaminated shellfish, finfish and  lobster.
Industries in this area were known to use  more AR 1242 than AR
1016orAR1254.'
  These  discrepancies  between principal  component  analysis,
Jaccard  measures  of similarity and visual  examination of  gas
chromatograms  demonstrate the difficulty associated with ascer-
taining  the nature of Aroclor* exposure. The techniques used
in this report are preliminary and experimental; it is hoped, how-
ever, that in vivo reference material will prove useful in delineat-
ing Aroclor®  exposure in people who live in proximity to waste
sites.
DISCLAIMER
  Use of trade names is for identification only and does not con-
stitute  endorsement by the Public Health Service  or by the U.S.
Department of Health and Human Services.
  246      ENDANGERMENT ASSESSMENT

-------
REFERENCES

 1.  Welty,  E.R.,  et at.,  "Assessing Exposure to Polychlorinated Bi-
    phenyls at Toxic Waste Sites." Unpublished Data, Centers for Disease
    Control.
 2.  National Priorities  List Data Summaries—The Mitre Corporation,
    Metrek Division, 1820 Dolley Madison Boulevard, McLean, VA, 1983.
 3.  Webb,  R.G. and McCall, A.C., "Quantitative PCB Standards for
    Electron Capture  Gas  Chromatography,"  Journal of Chroma to-
    graphic Science II, 1973, 366-373.
 4.  Brinkman, U.A. Th. and De Kok, A., "Production, properties and
    usage." InKimbrough, R.D. (Ed.), Halogenatedbiphenyls, terphenyl,
    naphthalenes,  dibenzodioxins and related products, Elsevier North-
    Holland Inc., New York, NY, 1980, 14-15.
 5.  CDC Laboratory Update 81-108 "Polychlorinated Biphenyl Determin-
    ation at Part-Ber-Billion Level in Serum."
 6.  Krull, I.S., "Recent Advances in PCB Analysis," Residue Reviews
    66, 1977, 185-201.
 7.  Robens, Hane and Anthony, H.D., "Polychlorinated Biphenyl Con-
    tamination of Feeder Cattle," J. Am.  Veterinary Medical Assoc. 117,
    1980,613-615.
 8.  Weaver,  G., "PCB  Contamination in  and Around New Bedford,
    Mass." Environ. Sci.  Technol. 18, 1984, 22A-27A.
 9.  Wold, S. and Sjorstrom, M. SIMCA: "A method for analyzing chem-
    ical data  in terms of  similarity and analogy."  In  Chemometrics:
    Theory and Its Application, Kowalski,  B.P., ed., ACS Symposium
    Series, 52. American Chemical Society, Washington, D.C., 1977.
10.  Massart,  D.L. and Kaufman,  L.,  The  Interpretation of Analytical
    Chemical Data by Use of Cluster Analysis. Wiley, New York, NY,
    1983.
11.  Johnson,  R.A. and Wichern, D.W., Applied Multivariate Statistical
    Analysis, Prentice-Hall, Englewood Cliff, NJ, 1982.
12.  Gordon,  A.D., Classification Methods for the Exploratory Analysis
    of Multivariate Data,  Chapman  and Hall, London: 1981,  19.
13.  Willett, L.B. and Hess, Jr. J.F., "Polychlorinated biphenyl residues
    in silos in the United States."  Residue Reviews,  F.A. Gunther, ed.,
    55, Springer-Verlag, New York,  NY, 1975, 135-147.
                                                                                        ENDANGERMENT ASSESSMENT
                                                              247

-------
EMERGENCY  PLANNING FOR ABANDONED WASTE  SITES
                                         TIMOTHY G.  PROTHERO
                                            Pittsburgh,  Pennsylvania
                                              JAMES FERGUSON
                                              WILLIAM  MARTIN
                                                      NIOSH
                                                 Cincinnati, Ohio
INTRODUCTION
  "Emergency" is a sudden, usually unexpected, change which re-
quires immediate corrective action. That definition is used in this
paper  and includes  terms  such  as  "accident",  "incident",
"event", and "happence", as synonyms for "emergency". There-
fore, in any hazardous material situation, any sudden change in
circumstances which prompts special corrective  actions  will be
termed an  "emergency," and those corrective  actions  will be
termed "emergency procedures". However, under this definition,
not all sudden changes are emergencies.
  The consequences of being unprepared to take proper correc-
tive actions for emergencies at abandoned hazardous sites can be
severe. Generally, most emergencies which lack advance prepara-
tions are characterized by the panic and confusion which  prevails
over all proper response activities. Personnel and visitors to the
area begin to react on impulse; some run away  from the emer-
gency,  while others  run toward it. The  response becomes hap-
hazard,  with several separate responding teams (if  any) reacting
without coordination and lacking a clear understanding of the
scope of the situation.
  To establish a coordinated, effective response to any emergency
situation, the OSC must determine the following within the first
few moments of the emergency:
•The type of emergency
•How to gather needed information
•How to distinguish rumor from fact
•The time requirements for the information gathering
•Time available before corrective actions are required
•Available response options
•Potential for wrong actions which could compound the situation
•How to determine the success or failure of the response action
  Time is the crucial element of any emergency response.  Time is
required for identifying the emergency, choosing the control  strat-
egy and executing the plans properly; yet the need remains for more
time than is available. One way to increase the amount  of time
available for response to an emergency is to use some time be-
fore any emergency to plan and rehearse a proper response.
TYPES OF EMERGENCIES

  To formulate  the best plans and procedures for  on-site emer-
gencies,  one should be aware of the types of emergencies to pre-
pare for. There are two major classes of possible  emergencies; (1)
one involving the workers and (2) one involving  the wastes. The
two types of situations will involve completely different response
actions, even if the emergency includes both workers and wastes.
The equipment,  procedures and training  for worker-related inci-
                                                    dents, which would require rescue and first aid, are totally different
                                                    than those required for waste-related actions, which would suggest
                                                    containerization procedures and fire-fighting. Therefore, it would
                                                    be advantageous to have two emergency action teams trained and
                                                    prepared for the differing aspects and types of emergencies. Key
                                                    considerations for each of these teams are:
                                                    •Something occurring to the worker
                                                      •accident
                                                      •physical injuries
                                                      •equipment failure
                                                      •medical problems
                                                      •chemical exposure
                                                    •Something occurring to the wastes
                                                      •leaks
                                                      •collapse of containers
                                                      •fire/explosions
                                                      •incompatible reactions
                                                      •release of toxic vapors
                                                      While many emergencies might be combinations of any of the in-
                                                    cidents mentioned, each aspect cited requires different procedures,
                                                    equipment and  training. Tables 1  and  2  contain examples of the
                                                    actions which are typically either "rescue" or "response".
                                                    WORKER-RELATED EMERGENCIES
                                                    Trips and Falls

                                                      Accidents, such as slips, trips or falls, are among the most com-
                                                    mon incidents and when no physical injuries result, are commonly
                                                    overlooked. In the situation where a worker trips and simply gets
                                                    up and proceeds as normal, few people would give such a minor in-
                                                    cident any notice. However, at hazardous waste  sites, it is impor-
                                                    tant to limit unnecessary contact with waste materials; falling to
                                                    the ground would place wastes directly on protective clothing which
                                                    does not have the thickness of the footwear. Consequently, perme-
                                                    ation breakthrough occurs sooner. A person who  falls while on-site
                                                    should at least proceed through a minimal decontamination pro-
                                                    cedure before  continuing with his  tasks or assignments. On-site
                                                    safety stations can include wash basins for such incidents.
                                                      Workers  who  trip  while wearing "level A" protective equip-
                                                    ment are not likely to be able to get up immediately even if they
                                                    are uninjured, due to the bulk of their equipment. Communica-
                                                    tions are necessary in such cases to allow others to know that there
                                                    is no injury so emergency rescue procedures are  not initiated un-
                                                    duly.
248
PERSONNEL SAFETY

-------
                           Table 1
                  Examples of Rescue Activities
           •Approach hazard scene carefully
           •Extricate trapped victim
           •Secure and remove unconscious victim
           •Communicate with conscious victim
           •Remove conscious victim from hazard area
           •Emergency decontamination
           •Emergency first aid
           •Transport victim to hospital
           •Reentry to obtain samples of wastes and/or air
           •Monitor rescue party for possible exposure
                           Table 2
                 Examples of Response Activities
            •Evacuate immediate area
            IFirefighting using
             •water
             •foam
             •dry sand or dirt
             •carbon dioxide
             •dry chemicals
            •Containment using
             •dikes
             •booms
             •diversion ditches
             •plug/patch methods
             •repack or overpack
            •Shut off feed
            •Isolation of emergency scene from other site areas
            •Monitor response crew for possible exposures
  Many accidents do involve physical injury, however, and every-
one should know what to do in these cases. The injured worker
must have some manner of communicating his injury to others and
his rescuers should know the proper first aid and rescue procedures
to follow in  a hazardous situation so they  do not unnecessarily
expose the victim to the chemical wastes or aggravate the injury.
  Physical injuries can occur as a result of an accident of the "slip,
trip or fall" variety or from some external physical blow resulting
from falling objects, explosions or other causes. The injury range
includes broken bones,  torn skin or damaged muscles.  The  care
required for  an emergency  involving physical injury is  well ex-
plained in many First Aid manuals; however, extra measures must
be taken when the victim is in a hazardous situation or area.

Equipment Failure
  Another purely physical problem which  can trigger  an emer-
gency  involving site  workers is equipment failure.  Abandoned
chemical waste sites have several widely varied types of hazards
against which the workers are protected by the wearing of certain
types of protective clothing  and respiratory equipment. The most
obvious type of possible equipment failure would involve the res-
piratory devices,  especially the SCBA. Any  failure of a SCBA is
going to place that worker in immediate danger.
  If a worker is wearing an SCBA,  then it is presumed that the ex-
ternal atmosphere is not fit to breathe and the safety areas are likely
too far to reach before panic sets in. Although respiratory equip-
ment is generally checked before each use, an emergency involving
air source failure is important enough to plan and prepare for, even
when considering the small chance of occurrence.
  Other more common, but less panic oriented situations of equip-
ment failure include: failure (by tear, penetration or permeation)
of the protective clothing, fogging  of face plates to zero visibility
and failure of air-purifying respiratory devices. In some cases, the
failure of equipment might not even be noticeable by the worker
using the equipment.

Medical Emergencies
  Medical problems are distinct possibilities for sudden emergen-
cies. Heat stress in the field, coupled with the heat capturing abil-
ities of the protective clothing, resistance to breathing of any res-
piratory protective devices and the work load which is increased by
the bulk of protective devices, can bring about heat stress and heat
stroke. The increased heat and work loads can aggravate existing
medical conditions. Even though  thorough medical screening  is
performed on the workers beforehand, the abnormally stressful
working  conditions can  aggravate minor conditions into  major
problems.
  Typically, one could expect to  discover stroke,  heart failure,
asthma, bronchitis, distortion of the senses (especially the sense of
balance)  or attacks  of claustrophobia. A background physical
which includes an exercise to place stress on the cardiovascular
system is preferable.
Chemical Exposure
  The last class of incidents which can affect site workers is ex-
posure to chemical agents. The  wide variety of chemical wastes,
where each waste is composed of numerous chemicals, means that
the toxic agent will not be identified in most cases of chemical ex-
posure. Most first aid responses to poisoning, as well as the medical
responses in a hospital, depend greatly on knowing the identity
of the toxic agent and the route of exposure.
  At abandoned hazardous waste sites, one would be lucky if he
could distinguish a chemical exposure from infection, such  as the
flu, let alone be able to positively identify the nature of the  chem-
ical to which the worker was exposed. Even in cases of direct con-
tact with wastes, such as  skin contact from splashing,  the waste
would be composed of so many different chemicals that the effects
and proper treatment would be difficult to determine. Therefore,
safety crews at remedial action sites should know beforehand what
procedures to follow in cases of known and suspected exposures.
WASTE-RELATED EMERGENCIES
Leaks

  Leaks and  releases  are  the most common occurrence at older
abandoned sites. Most such situations are looked at  as  common-
place and seldom as emergencies. Determining which leaks would
require an emergency response, and which ones  would be treated
as normal remedial responses, is entirely up to the discretion of the
OSC. Some of the factors to be considered include:
•Amount of material leaking (measured by volume/time and by
 volume-yet-to-be-released)
•Hazardous nature of the released material
•Impact on environment and neighboring population
•Time requirements for effects of pollutant to be felt
  Plug and patch kits, overpacks and other containment measures
should be available to control chemical releases, and staff should
be trained to use them.

  One  of the ways in which leakage occurs is the collapse  of the
containers; the drums or  tanks  deteriorate over time and  some-
times the actions of sampling are enough to cause the final collapse
of the containers. Therefore, all investigation visits should include
the preparation of sealing leaks and control of collapsing tanks or
stacks of drums. The preparations should include not only the con-
tainerization of spilt chemicals but  also any exposures and injuries
that might occur to workers within the immediate area.
Fire

  Of all  the hazardous properties  of chemical wastes,  the flam-
mable nature  of contained materials is recognized as one  of the
major risks in site work. Although a fire can be  easily ignited, ex-
tinguishing it may prove very difficult. Chemical fires are very diff-
icult to fight because different chemicals can require different fire-
                                                                                              PERSONNEL SAFETY
                                                         249

-------
 fighting techniques. While water will spread certain light solvents,
 other chemicals require foam or dirt to smother the fire. Some ma-
 terials cannot be smothered at all since they provide their own oxi-
 dizing source and are so classified. Some materials are water reac-
 tive, while some are air reactive. Thus, the wide variety of require-
 ments for chemical firefighting turns  waste site  fires into night-
 mares for everyone involved.
 Explosion

   Another aspect of waste site fires is explosion. Explosions can
 occur due to the unstable nature of the  chemical compounds
 (detonation) or due to the rapid combustion of a containerized ma-
 terial (deflagration).
   One type of emergency which can be triggered during remedial
 actions is a reaction due to the mixing of incompatible chemicals.
 Despite careful analysis of the various chemicals, the blending of
 incompatible substances can, and does, occur during site cleanup.
 The results of such reactions can range from toxic gas generation
 to fire and explosion.  Once blended together, there is  little that
 the OSC can do to halt the chemical reaction until it is complete.
 The response to severe incompatible reactions would be to treat the
 effects of any fire or explosion and to containerize as much of the
 released chemical as possible.
 Vapor Release
   During any leak, chemical release, incompatible reaction or fire,
 it is likely that toxic vapors may be released. Because most of those
 vapors are invisible, one should have a monitoring system in place
 to detect a variety of vapors and gases. For example, when bulk
                                                         wastes are solidified with lime, clouds of water vapor are usually
                                                         released due to the heat of reaction. It is important to remember
                                                         that many organic solvents have lower vapor pressures and boiling
                                                         points than water, and so entrained solvents would be released with
                                                         the water vapor.

                                                         CONCLUSIONS

                                                           The prevention of any site emergency is only one half of emer-
                                                         gency planning. As discussed, the other half involves recognizing
                                                         the possible hazards and formulating  the proper reactions  far
                                                         enough  in advance to  ensure smooth  and effective  emergency
                                                         responses.
                                                           The response to a waste-related emergency involves different
                                                         equipment (e.g., containment kits, firefighting equipment) than is
                                                         used in worker rescue operations (e.g., first aid and decontamina-
                                                         tion equipment). Any emergency plan should require different re-
                                                         sponse actions and equipment.


                                                         REFERENCES
                                                         I. Accident Prevention Manual for Industrial Operations, 7th ed., Na-
                                                           tional Safely Council, Chicago. IL, 1974.
                                                         2. Lcfevre, M.J., First Aid Manual for Chemical Accidents. Dowden
                                                           Hutchinson & Ross Inc., Stroudsburg, PA, 1980.
                                                         3. Planning Guide and Checklist for Hazardous Materials Contingency
                                                           Plans, Federal Emergency Management Agency, Washington, DC, July
                                                           1981.
                                                         4. Proctor, N.H. and Hughes, J.P., Chemical Hazards of the Workplace,
                                                           J.B. Lippincott Company, Philadelphia, PA, 1978.
250
PERSONNEL SAFETY

-------
      GUIDING  PRINCIPLES IN  THE  DESIGN OF  MEDICAL
            SURVEILLANCE PROGRAMS FOR HAZARDOUS
                                        WASTE WORKERS

                                        EDWIN C. HOLSTEIN, M.D.
                                     Environmental Sciences Laboratory
                                       Mount Sinai School of Medicine
                                             New York, New York
INTRODUCTION

  Evaluation and remediation of uncontrolled hazardous waste
sites is a rapidly growing industry. Employment,  already in the
thousands, is sure to grow. Improved techniques for dealing with
uncontrolled hazardous waste sites emerge almost daily, indicating
a steady improvement in the "product" of this industry.
Occupational Health Problems
  Like most new industries that  are vigorously expanding,  this
one has failed to make comparable progress in evaluating possible
occupational health hazards. Historical examples of occupational
health problems include  the asbestos debacle, in which asbestos
usage burgeoned without regard to evidence early in this century of
associated health hazards.
  The plastics industry expanded dramatically after World War II
but with scant attention  to occupational health. Later, angiosar-
comas among vinyl chloride reactor vessel workers were a sobering
reminder of past neglect.
  Off-shore oil exploration has  boomed  at times of petroleum
shortages. The very-deep diving required in this industry has pro-
duced anecdotal information suggesting serious long-term conse-
quences of chronic sub-clinical decompression illness but almost no
systematic study or standardized preventive measures.
  Numerous additional examples could be cited. The main point is
that new growth industries, beset by competitive pressures and in-
dustry "shake-out" and reluctant to miss out on growth opportun-
ities, frequently relegate occupational health to a very low priority.

Some Do It Right
  A different example of providing worker safety is found in the
genetic engineering industry. A decade ago, before this field had
been transformed from a  research activity to an industry, there was
great concern over potential health hazards. Many precautions and
restrictions were placed on genetic engineering laboratories, most
likely  because the potential risk was perceived to  extend beyond
employees to the general public.
  As a result, the evaluation of potential health hazards proceeded
hand-in-hand with the development of the industry. Moreover, the
required safeguards did  not significantly inhibit  the industry's
growth. When experience had shown that the actual health haz-
ards were not substantial, most of the precautions were rescinded.

UNCONTROLLED SITE HAZARDS
  Today, there is an alarming lack of information on occupational
health hazards related to uncontrolled hazardous waste sites. The
USEPA contracts for approximately 500 medical evaluations per
year of employees who work at such sites, and thousands more
are evaluated in state and industry programs. Yet there are almost
no published data on this topic. Nor are there data available on
actual toxic exposures of personnel, results of biological monitor-
ing or epidemiologic follow-up of employees.

USEPA Health Data
  Examinations of approximately 150 USEPA employees by the
Environmental Sciences Laboratory have not disclosed the presence
of any significant disorders attributable to exposure to hazardous
waste. A similar conclusion has been reached by Dr. Diana Ordin
on the basis of approximately 450 examinations of employees of the
New Jersey Department of Environmental Protection.
  According to interviewed  health and  safety officers of the
USEPA, there have been no hospital admissions for acute effects
of toxic exposures among USEPA hazardous waste workers. Only
one case requiring emergency room evaluation has occurred. There
have been no burns, only a few minor cases of trauma and four or
five incident reports. Only one complaint of a chronic skin disorder
has been lodged.
  Information of this sort is reassuring, but incomplete. It suggests
that acute  health hazards are uncommon among USEPA  hazard-
ous waste  employees. It also suggests that most subacute disease
developing over a period of weeks to a few years is uncommon.
Long-Latency Diseases

  On the other hand, several categories of disease still  cannot be
ruled out with the  information currently available. This includes
long-latency  diseases of many types, such as progressive organ
damage leading ultimately to  organ failure.  Phenacetin-induced
renal damage among analgesic users is an example from human
toxicologic experience. Gradual cumulative damage to the kidneys,
lungs, central and peripheral nervous system and liver are of par-
ticular concern. Long-latency carcinogenesis, of course, is another
category that cannot be ruled out with current information. Even
short-latency diseases  cannot  be ruled out if they are relatively
asymptomatic and undetected by routine medical tests. Reproduc-
tive failures, behavioral abnormalities and  immune dysfunction
are prominent in this category.

Cardiac Disease

  A final concern is the risk of sudden death due to cardiac disease.
The working conditions of employees in uncontrolled  hazardous
waste sites increase the risk greatly. Protective clothing and equip-
ment weigh as much as 55 Ib for USEPA employees, not includ-
ing loads of 50 to 100 Ib that may be carried on occasion
                                                                                       PERSONNEL SAFETY
                                                     251

-------
   The impermeability of protective clothing creates an enormous
 problem of heat accumulation, particularly in unshaded areas. This
 problem is compounded by the difficulty in providing safe drinking
 water in such sites. Respirators with two or more "stacked"  filter-
 ing cartridges greatly increase the work of breathing.
   For all these reasons, deaths due to cardiac disorders may  prove
 an even greater hazard to employees than toxic hazards.  To date,
 there do not appear to be any recorded incidents of such deaths,
 although heat exhaustion has been  quite common. This lack of
 problems is probably attributable to the fact that this new industry
 has mostly young employees.  When there are substantial  numbers
 of employees in their late forties and older, it seems almost inevit-
 able that there will be cardiac deaths regularly, unless current prac-
 tices are improved.

 CONCLUSIONS
   In summary, an examination  of what little information is  avail-
 able suggests that much work  remains to be done before there can
 be any complacency with regard to occupational health in the haz-
 ardous waste industry. The major areas requiring further work are:
 •Characterization of the working environment (inside  and outside
                                                          of protective clothing);
                                                         •Biological monitoring of toxic substances and/or their metab-
                                                          olites;
                                                         •Medical surveillance carried out by physicians knowledgeable with
                                                          occupational health problems and practices;
                                                         •Systematic, centralized review and analysis of findings; and
                                                         •Long-term epidemiological follow-up and analysis.
                                                           With regard to medical surveillance, current practices require
                                                         supplementation  to test for insidious disease processes with  long
                                                         latencies. In addition, cardiac  risk factors must be evaluated  in
                                                         greater depth.
                                                           Such recommendations may appear excessive in a setting where
                                                         there have been few documented health problems to date. How-
                                                         ever, one  must remember problems (or lack thereof)  that have
                                                         occurred in other developing industries to appreciate the wisdom of
                                                         this approach. It is likely that, after a number of years of such
                                                         effort have been accomplished, only a few actual hazards will have
                                                         been identified and numerous potential hazards  found to be non-
                                                         existent. At that point, but only then, the medical surveillance pro-
                                                         gram can be cut back to the efficient, focused effort seen in mature
                                                         industries.
252
PERSONNEL SAFETY

-------
        ASSESSMENT OF HUMAN HEALTH  EFFECTS FROM
             EXPOSURE TO  ENVIRONMENTAL TOXICANTS

                                       KAREN K. STEINBERG, Ph.D.
                                        LARRY L. NEEDHAM, Ph.D.
                                         JOHN T. BERNERT, Ph.D.
                                           JAMES MYRICK, Ph.D.
                                           DAVID D. BAYSE, Ph.D.
                                          Centers for Disease Control
                                                 Atlanta, Georgia
INTRODUCTION
  On August 10, 1984, the House of Representatives voted on
several amendments to House Bill HR5640, the "Superfund Ex-
pansion and Protection Act", as well as on the bill itself. One of
the amendments called for the allocation of  12% of the  $10.2
billion for compensation for medical care and lost wages to persons
who  had   experienced  adverse   health  effects  which  were
"reasonably likely" to have been caused by exposure to hazardous
wastes from one of the hazardous waste sites identified by the
USEPA. Although  the bill was  passed, the amendment was
defeated. One argument in opposition to the compensation amend-
ment  was that in the majority of cases the scientific means to
establish whether an effect is the result of a putative exposure to a
toxicant are not available.
  A primary objective of the Clinical Chemistry Division of the
Center for Environmental Health, Centers for Disease Control, is
to develop methods to identify adverse health effects and to deter-
mine  whether these  effects are "reasonably  likely" to have been
caused by hazardous wastes.
  Assessment of health effects of hazardous  wastes  must  begin
with a definition of what it is that is sought. What is an adverse
health effect?  Russel  Sherwin of the University of Southern
California School of Medicine has defined an adverse health effect
as "the causation, promotion, facilitation and/or exacerbation of a
structural or functional abnormality, with the implication that the
abnormality produced has the potential of lowering the quality of
life, causing a disabling illness or leading to premature death"1;
stated briefly, the causation or abetting of an abnormality that can
result in altered quality of life, disease or death. This definition
stresses the very early stages of disease, the point at which there is
potential for disease.
STAGES IN DEVELOPMENT
OF ADVERSE EFFECTS
  The progression from health to end points such as overt disease
and death constitute a spectrum that can include an adaptive state,
subclinical disease and the point at which adverse health effects are
most commonly detected, symptomatic disease or death (Fig. 1).
The insult, that is confrontation with the hazardous substance or its
metabolite at the cellular level, can, and sometimes must be, pre-
sent in the stages beyond the adaptive stage.  Resolution of the in-
sult can also occur at many points in this process.

Adaptation

  Dr. Sherwin's definition might  be qualified  by saying that an
event which is not an abnormality, but which may obtain the same
results, is an adverse health effect. Some of the body's responses to
stressful environmental factors, such as chemicals, are adaptive,
helpful responses aimed at restoring a biochemical or physiological
balance. These adaptive states often lead to the resolution of an in-
sult without the host knowing that they have been challenged.
  Although these responses are not abnormalities, there are cases
in which this mechanism of adaptation can produce chemicals more
dangerous than those with which  the body was originally con-
fronted. Cellular enzymes called mixed function oxidases (MFO)
are bound to cell membranes that are  called the endoplasmic
reticulum in the intact cell and microsomes when isolated from the
cell. The MFO metabolize endogenous compounds, such as fatty
acids and steroid hormones, and lipophilic exogenous compounds,
such as polychlorinated hydrocarbons and other xenobiotics, to
more polar compounds that  can be excreted or conjugated with
other   compounds,  such  as glucuronic  acid,  sulfates  or
gluthathione, and then excreted. Depending on  the life-span of
these polar metabolites of the MFO, they can bind to cellular
macromolecules such as DNA, RNA or protein or initiate perox-
idation of membrane lipids before they are further metabolized by
conjugating enzymes. This formation of  xenobiotic adducts with
cellular  macromolecules can ultimately  result in  cancer, birth
defects, mutagenesis or other organ or system malfunctions.2
  Enzymes which conjugate  endogenous  compounds, such as
bilirubin and steroid hormones, and exogenous compounds, such
as toxicants or their metabolites, are present in the endoplasmic
reticulum with the MFO or in the cytosol.  Enzymes of the MFO
system and some enzymes of conjugation (such as glucronyltrans-
ferase and glutathione-S-transferase) are  inducible: they are syn-
thesized in response to the presence of their substrate, the com-
pound  that  they metabolize. This property of  induction  by
xenobiotics and the subsequent increase of products of these  en-
zyme pathways is being investigated as supportive  evidence for the
presence of toxicant. D-glucaric acid is a urinary end-product of an
inducible conjugation pathway, the glucuronidation/deglucuroni-
dation  pathway. In  one  study,  elevated concentrations  of
D-glucaric acid in the  urine of children living in  an area con-
taminated with  2,3,7,8-tetrachlorodibenzodioxin (TCDD)  cor-
related with the presence of chloracne.3
    aun
    X
         ADAPTIVE STATE
                     r*
                        SUBCLINICAL DISEASE — ^-DISEASE
                                                   DEATH
                    resolution / Tewogenesis
                          I Carcinogenesis
                          V Oil Death
Birth Defects
Cancer
Other Disease
e.g. hepatitis
                         Figure 1
        Spectrum of States Resulting from Toxic Exposures
                                                                                      PERSONNEL SAFETY
                                                     253

-------
   If there is evidence that enzymes are being induced, is there also
 evidence of an adverse health effect? In some cases, the answer is
 yes and in others no. The primary determinants of toxicity are the
 amount and  life-span of the intermediate.  Other factors such as
 original dose of toxicant, route of entry and  other host factors, not
 all of which are well understood, determine  whether the metabolic
 product of enzyme pathways will be harmful.
   Methods are needed to identify the adaptive phase and to be used
 in prospective studies  to determine in which  cases the adaptive
 phase signifies an adverse health effect. If  induction of enzymes
 results in adduct formation with the toxicant or its metabolite and
 the process can  be detected, intervention may be possible by ad-
 ministering  chemical trapping  agents that prevent  binding to
 cellular macromolecules. Such intervention  has been successful in
 ameliorating the toxic  effects  of poisoning with  drugs such as
 acetaminophen.4
   For the present, adverse effects must be detected beyond the
 adaptive state. Ideally,  those at risk for adverse health effects
 should be identified using either actual body burden of toxicant or
 a surrogate measure of body burden. Within this selected group, a
 subgroup should be identified. In this case, an event that is closely
 associated with disease, but which occurs before subclinical disease,
 i.e., before there are biological changes in the  cell causing altered
 cell function or cell death. Examples of such events are the binding
 of benzy(a)pyrene, or its metabolite, to DNA  and  the binding of
 TCDD to a receptor protein in the cytosol.5'6 If these compounds
 could be identified bound to cellular macromolecules in accessible
 tissues (such as blood cells), placenta or scrapings of the mucous
 membranes, their detection in population studies would be feasible.
 Sensitive immunoassays which identify DNA-toxicant adducts in
 femtomolar concentrations have been developed and are of poten-
 tial use in predicting risk for adverse health effects.1
 Subclinical Disease

   Subclinical disease is  the point at which disease is  an almost
 assured result without intervention. Biological and/or biochemical
 abnormalities and  perhaps cell death are  present at this  stage.
 Symptoms are not apparent at  this point.  Biochemical markers,
 such as slight elevations in serum concentrations of enzymes due to
 altered cell membrane permeability, are being sought to identify
 this early stage  of disease. A  combination of measurements of
 alkaline phosphatase, bile acids and indocyanin green clearance can
 predict the presence of  histological lesions in lungs of apparently
 asymptomatic persons exposed occupationally to vinyl monomers.'
 Host Factors

   The presence of a toxicant within  a cell is evidence of exposure
 but does  not always predict disease.  Factors which  are host-
 specific, which cause one person to be susceptible to the hazards of
 environmental chemicals and others to be resistant, will need to be
 evaluated to determine actual risk after exposure is documented by
 body burden or a biological effect. Host resistance depends largely
 on genetic factors. The innate competence of the immune system,
 the types and amount of enzymes produced and DNA repair when
 toxicant-adduct   formation  occurs,  are  genetically determined.
 Confounding environmental factors, such as poor nutrition or in-
 fection with hepatitus B  virus, increase susceptibility.

 METHODS FOR MONITORING
 HEALTH EFFECTS

  Assays for assessment of adverse  health  effects  have been
organized into organ or  organ-system categories by the Clinical
Chemistry Division of the Centers  for Disease  Control. These
categories include the liver, the kidney, the cardiovascular system
and the immune system. It is true that the nervous and reproductive
systems  are  targets for environmental  toxicants.  At  present,
however, there are no biochemical tests on accessible body fluid or
tissue more sensitive than the methods used  by  the  clinician in
evaluating neurological  function. There are certain  assays that
would give limited  information  about the  reproductive system,
                                                           such as chromosome rearrangement or DNA-toxicant adduct for-
                                                           mation in sperm. These systems may be included at a later date.

                                                           Liver Profile
                                                             The liver profile consists of tests that were selected to reflect in-
                                                           ductive or adaptive changes, damage to liver cells (hepatocellular
                                                           damage) or impaired flow of bile (cholestatis), since bile flow is a
                                                           major function of the liver (Table 1). The compounds with the suf-
                                                           fix "asc" are enzymes, proteins which regulate the speed of reac-
                                                           tions in the body. Beta-glucuronidase, gammaglutamyl transferase,
                                                           alkaline phosphatase and glutathione-S-transferase are inducible
                                                           enzymes and are included with the hope of detecting this early stage
                                                           of induction. These four enzymes  are also released in cell destruc-
                                                           tion and can be  an indication  of subclinical or overt disease.

                                                                                      Table 1
                                                                                    Liver Profile
                                                                        Inductive or Adaptive
                                                                          Urinary 0 Glucaric Acid
                                                                          Serum (j.immaglutamvl Tramlerai*
                                                                          Scrum Beta Glucuro«iida«
                                                                          Serum Glutathione S Transferase (GST)
                                                                          Serum Join1 Bilirubm IDecreated)
                                                                          Serum Alkaline Phosprujui* (5 Nucleotidase
                                                                           or Alkaline Phosphatate Isoeniymes!
                                                                          Urinary Mercaptunc Acids
                                                                          Ratio ol 0«idi/ed to Reduced Glutathione in
                                                                            RSCiand WBCt

                                                                        Hepatocellular Damage
                                                                          Serum Alanme Aminotransferase
                                                                          Serum Aspartate Aminotransferase
                                                                          Serum G S-T
                                                                          Serum Total and Direct Bihrubm
                                                                          Serum Ornithine Cartumovi
                                                                           Translerase (OCTl
                                                                          Serum Lactate Dehydrogenase
                                                                          Serum Sorbitol Dehydrogenise ISDHI
                                                                          Albumin/Globulin


                                                                       Cholestasis {Impaired Formation of Canalicutar
                                                                       Bile or  Its E.ill
                                                                          Serum Bile Acids
                                                                          Serum Total and Direct Bilirubin
                                                                          Serum Alkaline Ptiosprvataw
                                                                          Serum Alanme Aminopeptidase

                                                                       Additional Tests
                                                                          RBC Count and Indices
                                                            Although  differentiating  induction  from release  due to  cell
                                                          destruction can be difficult, the presence of enzymes such as aspar-
                                                          tate aminotransferase and alanine aminotransferase, which are not
                                                          inducible but which are usually elevated due to cell  destruction,
                                                          should help make this distinction.
                                                            With the exceptions of sorbitol dehydrogenase and ornithine car-
                                                          bamoyl transferase, these  enzymes are not unique to liver cells.
                                                          However, the liver is relatively enriched in  most of these enzymes
                                                          compared  to other  tissues. Alkaline  phosphatase  and  lactate
                                                          dehydrogenase are physically separated into components called
                                                          isoenzymes which represent species coming from different tissues.
                                                          5'nucleotidase is  a liver specific form of alkaline phosphatase that
                                                          may help identify the source of the alkaline phosphatase that is
                                                          elevated  in serum.
                                                            Although these enzymes may definitively reveal neither which
                                                          process, adaptation or destruction is occurring,  nor in which tissue
                                                          that process  is occurring,  when  taken  together in the context of
                                                          other tests, they can provide evidence regarding the process and the
                                                          site.
                                                            Cholestasis is a condition  in which there is decreased bile flow
                                                          and can  result at the level  of the hepatocyte following xenobiotic
                                                          exposure. Enzymes that are present in serum in increased concen-
                                                          tration due to cholestasis are alkaline phosphatase, 5'nucleotidase,
                                                          alanine aminopeptidase and gammaglutamyl transferase.
254
PERSONNEL SAFETY

-------
  As mentioned above, D-glucaric acid is an end product of the
glucuronidation/deglucuronidation pathway  and is  excreted  in
urine. Elevated urinary D-glucaric acid concentrations are believed
to be evidence of the liver's attempt to metabolize foreign or en-
dogenous compounds.
  Bilirubin is a  breakdown  product  of hemoglobin which is
metabolized and excreted by the liver in bile. Serum bilirubin con-
centrations are an indicator of liver function.
  Mercapturic acids in urine are the end result of the binding of the
highly reactive metabolites of xenobiotics to glutathione, a com-
pound found in the cytosol of cells which acts as a trapping agent
for these reactive compounds.  When cellular glutathione stores are
depleted, adduct formation can readily occur.
  There is overlap in the classification of tests listed here. For ex-
ample, when bilirubin is increased, there is usually either damage to
liver cells and/or  blockage of bile  flow. When  bilirubin is de-
creased, enzymes of conjugation have been induced and are causing
bilirubin (an endogenous compound) as well as the toxicant (an ex-
ogenous compound) to be excreted with resulting lower blood con-
centrations.  Bilirubin,  therefore, is  found  in  more  than one
category.
   Four of these analytes have  been quantified in male rats ad-
ministered a single  oral dose of 500 mg/kg of polybrominated
biphenyls (PBBs)  by gavage. Mean alanine aminopeptidase and
sorbitol  dehydrogenase  concentrations  from dosed  rates  were
significantly higher than those of the controls (Fig. 2). Gammaglu-
tamyl transferase concentrations were higher in PBB-dosed rates
than in controls, but not significantly so. Mean beta-glucuronidase
concentration from dosed rats was significantly lower than that of
controls. Mean bile acid concentration was significantly higher in
test animals.  Interpretation of these data will follow completion of
the other assays in the liver profile.
   In a population at risk  for exposure to TCDD, mean  aklaline
phosphatase, beta-glucuronidase and  lactate dehydrogenase con-
centrations were higher in the total  TCDD-exposed population
than in a reference group matched for mean age, sex, alcohol and
drugs (Table 2).' Mean gammaglutamyl transferase concentration
distinguished persons with high blood PBB concentrations ( >150
fig/I)' from those with low blood PBB concentrations (< 5 jtg/0-7
Additional animal and human studies are planned or are underway
to determine which tests, or groups of tests, in this profile give per-
tinent information regarding exposure and health effects.
      Serum alanine aminopeptidase, (AAP), sorbitol dehydrogenase (SDH),
      gamma-glutamyl transferase, (GGTI, beta-glucuronidase, (beta-glu), bile
      acids, (BA) on rats dosed with 500 mg/Kg PBB (left) and controls
      (right). Mean values ± 2SD in figure, mean values ± ISO above figures.
      Number of rats in parentheses
         AAP'
                    SDH'
                             GGT
                                       beta-glu"
                                                  BA'
       91.9    78.7    87.8 60.1  8.4   6.8   1.86  2.24  12.7  6.77
      ±22.8   ± 6.7   ±25  ±31   ±4.5  ±4.8  ±0.38 ±0.38 " 7.6 ±6.8
       (10)    (10)    (9)  116)  (7)   (17)  (10)  (10)   (91   (131
130
120
•110
•100
•90 <
>80
70
60
SO





(



•40
30
•20
10





t










100
{



•50





I


(




•10






\








• 20




•10
4




















6 0

5 0

4.0

3.0

.2.0,

i.o-'










t

















-30



•20


(
•10











1 i


      'Differences in means significant with P <0.05.
                            Figure 2
Kidney Profile
  The kidney profile includes four enzymes that are often elevated
in urine after acute exposure to toxicants because of sloughing of
cells lining the tubules of the kidneys (Table 3). These enzymes are
gammaglutamyl transferase,  lactate dehydrogenase, n'acetylglu-
cosaminidase and alanine aminopeptidase. All of these enzymes are
found in many tissues other than kidney, but elevated urinary con-
centrations can help to confirm an acute toxic insult. Many other
enzymes have been  demonstrated to be elevated in urine due to
acute exposure. Four  enzymes were selected; they were  enzymes
consistently reported in the literature as being useful.9'10
  A relatively  new,  high resolution technique called two dimen-
sional electrophoresis is used to  identify and quantify proteins in
                                                               Table 2

                                                  Bilirubin, gammaglutamyl transferase (GGT}, beta-glucuronidasa,
                                                  alanine aminotransferase (ALT), aspartate aminotransferase (AST),
                                                  lactate dehydrogenase (LDH), alkaline phosphatase (Alk. Phos.),
                                                  and triglycerides in populations exposed to PBBs and TCDD.

Bilirubin Direct Img/dl)
(0.0-0.51 1
Total Img/dl)
(0.1 - 1.3)
GGT ImU/L
(0-60, male; 0-45, female)
(J-glucuronidase (u/L)§
ALT ISGPT) (mU/ml)
(5-401
AST (SCOT) ImU/mL)
(10-401
LDH (mU/mL)
190-255)
Alk. Phos. ImU/mLI
120-95)
Triglycerides (mg/dl)
(10-150)
'Mean values ±

High PBB
N-18
0 1710.2'
0.2T
0.51 ±0.2
0.4
25*26
16
2.18±0.8
2.00
16±8
15
21119
19
173±50
154
46±11
44
103148
95
ISO in u/L
Michigan
Low PBB
N-84
0.210.2
0.2
0.5410.2
0.5
15121
5
2.2410.94
2.04
18110
16
21119
18
161144
155
52115
49
123162
120


All
N=114
0.1910.1
0.2
0.5410.2
0.5
16122
5
2.2410.95
2.03
1819
16
21118
116
163145
154
51115
49
120*60
116

•"•Median values in u/L

High Risk
N=50
0.210.13
0.2
0.5810.3
0.5
16121
9
2.7811.13
2.56
16±15
11
1918
18
172131
175
63124
62
107155
90
; Reference
Missouri
Low Risk
N=27
0.2710.17
0.2
0.6810.35
0.6
18+24
5
2.4311.03
2.27
18113
13
1717
15
180125
172
64131
56
105164
77
limits

All
N=77
0.2310.15
0.2
0.6110.3
0.6
17122
5
2.66H.1
2.37
17±14
12
1917
18
175132
170
63126
57
107157
88

Reference
N=46
0.2210.1
0.2
0.4510.2
0.35
1719
16
1.9110.82§
1.78
17110
16
1915
18
151128
144
47±12
46
94144
88

§ For the Reference Group : N=71
                                                                                                   PERSONNEL SAFETY
                                                            255

-------
                             Table 3
                          Kidney Profile
                 Gamma Glutamyl Transferase
                 Lactate Dehydrogenase
                 N'acetylglucosaminidase
                 Alanine Aminopeptidase
                 Two dimensional Electrophoresis
                    of Urinary Proteins
 body fluids; in this case, urine.  Briefly,  the method includes con-
 centrating urine specimens and applying the concentrate to a cylin-
 drical gel in a tube (Fig. 3). An electrical current is applied and pro-
 teins are separated according to their charge. This is the first dimen-
 sion. The tube gels are then placed across the top of a flat, slab gel
 with increasingly  smaller pore sizes. An  electrical current is again
 applied and the proteins separate according to  molecular weight.
 This is the second dimension. The slab gels are then stained with a
 sensitive  silver  stain and a pattern results.  That pattern is  inter-
 preted using  a computer-assisted imager. Rats  dosed orally with
 PBBs as described  above  had  protein  patterns  which  were
 significantly  different from  those of controls (Figs.  4 and 5).
 Studies are planned to evaluate urinary protein patterns in human
 populations which  have  been accidentally  exposed to toxicants.
 Generally, urinary enzymes are useful in detecting early  injury to
 the  renal tubules  in acute exposures.  Chronic toxicity more often
 results in nephrotic syndrome which is evidenced by increased con-
 centrations of larger molecular weight proteins in urine.''
       Application  o*
     protein  to  tub*
     gel for teparation
     bated on charge
  Application  of
charge and  vpa
ration o* protein
 Tube gel pi*ced
acrou top ol tl*t>
gel and protein a/e
separated accord
ing to rnolecuUr
weight
                             Figure 3
            Two Dimensional Electrophoresii of Proteins
Cardiovascular Profile

  The cardiovascular profile is a profile of blood tipids (fats) and
proteins that are closely associated with them in the blood (Table
4). Because lipids are insoluble in the blood, they are transported as
complexes with proteins which are soluble. Concentrations of these
lipids and proteins have been used to determine risk  for develop-
ment of cardiovascular disease, hence the category cardiovascular


                           Figure 4
                        PBB-Dosed Rats

256       PERSONNEL SAFETY
                            Figure 5
                          Control Rats

-------
                            Table 4
                     Cardiovascular Profile
                            Table 6
           Effect of Polybrominated Bipbenyls on Hepatic
                Microsomal Fatty Acid Composition
             Cholesterol
             Trkjlycerides
             HDL-Cholesterol
             Apolipoprotein A-1
             Apolipoprotein A-2
             Cholesterolester Fatty Acid Composition
             Phospholipid Fatty Acid Composition
system. Blood concentrations of cholesterol, triglycerides and high
density lipoprotein-cholesterol (HDL-cholesterol) are known to be
altered by certain toxicants, possibly as an indirect result of liver
dysfunction.12 Alterations in concentrations of two lipid-associated
proteins, apolipoproteins A-I and A-II, are being investigated in
toxic exposure.
  Fatty acids are long chain hydrocarbons with a carboxyl group
through which they may covalently bind to other compounds (such
as phosphoglycerides), forming phospholipids (important com-
ponents of cell membranes).
  Rats given a single oral dose of 500 mg/kg PBBs were sacrificed
at intervals  of  from  1  to 8  weeks.  They  had  alterations in
microsomal membrane phospholipids when compared to controls
(Table 5). Mean liver weight in the PBB-dosed rates was higher
than that of controls (Table 6) which is consistent with the pro-
liferation of the  lipids and proteins of the encoplasmic reticulum
during an adaptive phase.13
Immune Profile

  The immune system is a complex interaction of many types of
white  blood cells, tissues, hormones and other components. It is
difficult to pinpoint an abnormality in the immune system without
extensive testing. There are basically two types of immunity; cell-
mediated and antibody-mediated or  humoral immunity. These two
types of immunity can be assessed  separately, although they are
often  interdependent. Cell-mediated  immunity is  dependent on
lymphocytes  (white  blood  cells) called T-cells,  and  antibody-
mediated immunity is dependent on lymphocytes called B-cells.14
  Adverse effects of toxicant exposure include suppression of cell-
mediated immunity, which protects against development of cancer

                            Table 5
                        Immune Profile
         White Blood Cell Count with Differential Smear
              T-cell Profile
                    T-helper
                    T-suppressor
                    Null Cells
              T-cell Function (Proliferation Tests)
                    Phytohemmagglutinin
                    Concanavalin A
                    Pokeweek Mitogen
              B-cell Profile
                    Serum Immunoglobulin Concentrations
                                 IgG
                                 IgA
                                 IgM
                    Skin Tests
                         Tetanus Toxoid
                         Diphtheria
                         Streptococcus
                         Tuberculin
                         Candida
                         Trichophytin
                         Proteus
Fatty Acid

14:0
16:0
16:1 (r>-7)
18:0
18:1 (n-9)
18:1 ln-7|
18:2 (n-6)
20:3 (n-6)
20:4 ln-6)
22:4 (n-61
22:5 (n-61
22:5(n-3l
22:6 (n-3)



PBB-Dosed (X±SEM|Wt%
(n= 15)
0.31
15.40
1.49
26.93
11.84
2.94
9.54
0.86
25.87
0.43
0.46
0.44
3.01
.02
.18
.14
.59
.82
.17
.31
.04
.5V
.04 (NSI
.06INSI
.03 (NSI
.17
P<.05
NS = not significant
All others: p < .01
Control (X ± SEMI
In = 8!
0.59
23.50
3.32
19.07
7.08
3.99
12.07
1.16
23.83
0.36
0.35
0.51
4.13
.05
.87
.25
.31
.16
.29
.45
.11
.60
.06
.05
.03
.35



      N is the number of microsomal preparations. Each preparation was obtained
    from two pooled livers. All of the exposure periods were combined for this com-
    parison. Trivial names for the first nine acids given in this Table are: myristate
    (14:0), palmitate (16:0), palmitoleate (16:1, n-71, stearate (18:0), oleate 118:1,
    n—91,  cis-vaccenate (18:1, n—7), linoleate (18:2, n—6), dihomo-7 linolenate
    (20:3, n-6) and arachidonate (20:4, n-6).
                            Table 7
           Hepatic Weight Response to a Single Oral Dose
                   of Polybrominated Biphenyls
Weeks

1

2

4

8

Control15

Liver Weight3
(9)
22.13± 1.20
(8)
22.32+ 1.37
(8)
23.35 ± 0.89
(81
26.13 + 1.35
(7)
14.81 ± 1.37
(4)
% of Body Weight

4.47 ± 0.21
(8)
4.51 ± 0.20
(8)
4.70 + 0.12
(8)
4.9410.21
(7)
2.88 + 0.80
(4)
       aAII values are x ± SEM; the number of animals for each group is
        given in parentheses.

       ''Four-week data only.
and viral infections, and may include abnormalities in antibody-
mediated immunity, which protects us from bacterial infection.15-16
In testing for immune function, total numbers of lymphocytes are
determined and the ratios of subgroups of T-cells (T-helper and
T-suppressor cells) are determined (Table 7). These cells are also
tested  to determine if they are  functional  by exposing them to
foreign compounds called mitogens, which stimulate cell division.
This method is called mitogen stimulation testing. There are many
other components of cell-mediated immunity, but T-cell function is
the principal component for understanding this function of the im-
mune system. B-cell immunity can be evaluated by determining an-
tibody concentrations  in serum.
CONCLUSIONS

  A wide variety of assays that may help assess health effects by
determining whether a person or population has a toxicant burden,
whether an event has occurred that increases their risk for disease
(the  binding   of  toxicant  or its   metabolite  to  cellular
macromolecules) or whether there is a biochemical or functional
abnormality that may be a result of exposure to a toxicant has been
enumerated. This group of assays will  be evaluated for use  in
                                                                                                 PERSONNEL SAFETY
                                                           257

-------
 assessing  health  effects.  Statistical  methods  enabling  us  to
 recognize patterns of test results which correlate with body burden
 of toxicant or inapparent histological changes are in place to help
 evaluate these assays. Present technology allows determination of
 the concentration of a toxicant in the environment and, in many
 cases, the body burden of the toxicant or its  metabolite. The link
 between these events and overt disease is elusive. A combination of
 controlled prospective animal, clinical and epidemiological studies
 will establish these relationships. Until these studies are complete,
 relative risk for development of disease, time sequence of exposure
 and development of disease, knowledge of a biological mechanism
 that could  explain  the  relationship  (as  is  the case  for  high-
 cholesterol  diets  and heart  disease) and  other epidemiological
 criteria are evidence of cause and effect relationships.
 REFERENCES

  1.  Sherwin, R. "What is an Adverse Health Effect," Environ. Health
     Perspectives 52; 1983, 177-182.
  2.  Miller, E.G., "Some Current Perspectives on Chemical Carcinogcne-
     sis in Humans and Experimental Animals: Presidential Address. Can-
     cer Research 38; 1978, 1479-1496.
  3.  Ideo, G., Bellati, G., Bellobuono, A.,Mocarelli, P., Marocchi, A. and
     Brambilla,  P., "Increased  Urinary  D-glucaric  Acid Excretion  by
     Children Living  in an Area  Polluted with Tetrachlorodibenzopara-
     dioxin (TCDD)," Clin. Chim. Acta 120; 273-283.
  4.  Mitchell, J.R., Thorgeirsson, S.S., Potter, W.Z.,  JoUow,  D.J. and
     Reiser, H., "Acetaminophen-induced Hepatic Injury. Protective Role
     of Glutathione in Man and Rationale  for Therapy," Clin. Pharm.
     Ther., 16, 1974, 676-684.
  5.  Perera,  F.P.  and Weinstein, I.B.,  "Molecular Epidemiology  and
     Carcinogen-DNA Adduct Detection:  New Approaches to Studies of
     Human Cancer Causation," /. Chronic Disease, 35, 1982, 581-600.
                                                             6.  Poland, A. and Knutson, J.C., "2,3,7,8-tetrachlorodibenzodioxin and
                                                                Related Halogenated Aromatic  Hydrocarbons: Examination of the
                                                                Mechanism of Toxicity," Ann. Rev, Pharmacol. Toxicol., 22, 1982,
                                                                514-517.
                                                             7.  Steinberg, K.K., Smith, B.F. and Steindel, S.J., "Activity of three
                                                                Serum Enzymes after Exposure to Environmental Toxicants," Manu-
                                                                script in preparation.
                                                             8.  Needham, L.L., Burse, V.W. and Price, H.A. "Temperature Pro-
                                                                grammed  Gas Chromatographic Determination  of Polychlorinated
                                                                and Polybrominated Byphenyls  in Serum," J. Assc. Off- Analyl.
                                                                Chem., 1981, 1131-1137.
                                                             9.  Price, R.G., "Urinary Enzymes," Nephrotoxicily and Renal Disease.
                                                                Toxicol.. 23,  1977. 99-134.
                                                            10.  Ngaha, E.O. and Plummer, D.T., "Toxic Renal Damage: Changes in
                                                                Enzyme Levels," Chemical Medicine, 18, 1977, 71-79.
                                                            11.  Kluwe, W.M.,  "Renal Function Tests as Indicators of Kidney In-
                                                                jury in Subacute Toxicity Studies," Toxicol. Appl. Pharmacol.. 57,
                                                                1981, 414-424.
                                                            12.  Lambercht, L.K.,  Barsotti, D.A. and Allen, J.R., "Response of Non-
                                                                human Primates to a Polybrominated Biphenyl Mixture," Environ.
                                                                Health Perspect.. 23. 1978. 139-145.
                                                            13.  Bernert, J.T.  and  Groce, D.f. and Kimbrough, R., "Long-Term Ef-
                                                                fects of a Single Oral Dose of Biphenyls on Serum and Liver Lipid
                                                                in Rats."  Toxicol. Appl. Pharmacol.. 68, 1983, 424-433.
                                                            14.  Greaves. M.F., Owen.  J.J.T. and Ruff, M.C., "T and B Lympho-
                                                                cytes: Origins,  Properties  and Roles in Immune  Responses," Ex-
                                                                cerpta. Medico., Amsterdam, 1973.
                                                            15.  McConnell, E.E.,  Moore. J.A.  and Dalgard. D.W., "Toxicity of
                                                                2,3,7,8-tetrachlorodibenzo-pdioxin   in  Rhesus  Monkeys  (Macaca
                                                                mulatta) Following a Single Oral Dose," Toxicol. Appl. Pharmacol.,
                                                                43, 1978, 175-187.
                                                            16.  Vos, J.G., Faith, R.E. and  Luster, M.I.. "Immune Alterations in
                                                                Halogenated  Biphenyls, Terphenyb, Naphthalenes, Dibenzodioxins
                                                                and Related Products. R. Kimbrough, ed.. Ebevier/North Holland,
                                                                1980, 241-266.
258
PERSONNEL SAFETY

-------
                         THE MEDICAL SURVEILLANCE OF
                            HAZARDOUS WASTE WORKERS

                                              FRANK L. MITCHELL
                                            Centers for Disease Control
                                                  Atlanta, Georgia
INTRODUCTION
  The publicity given to Love Canal, dioxin contamination and
chemical spills in emergency situations, has focused attention on
the dangers of hazardous wastes. All too common are pictures of
people leaving, or wondering whether to leave, their homes while
workers in "moon suits" use exotic testing equipment to evaluate
the hazard or cleanup a site.
  These situations have brought new challenges to health profes-
sionals because of the unconventional types of exposures, both ac-
tual and potential, and the inability to clearly define the expected
effects. This  complex challenge is particularly important to  the
occupational health professional who must deal with the workers
that are cleaning up these waste sites.
  In this paper, the  author describes some of the problems in-
volved, and addresses the provision of medical surveillance to these
workers.
PROBLEMS FACED
  Even those trained in occupational or environmental medicine
are usually not prepared for the complexity of the problems at these
sites. A typical abandoned waste site will be:
    In a rural, probably wooded, area. Thus, it is out  of the
  way and out of sight of law enforcement officials. There are
  hundreds, perhaps thousands, of drums in varying states of
  deterioration. There may be a few labels  on some  of the
  drums that list contents but, because of their condition and
  age, there is no way of knowing whether the labels have any
  relation to the current contents. Some leaks appear to be oil
  and may indicate the presence of PCBs. Since many  drums
  seem to have some connection with pesticides, there may also
  be dioxins present.
    There are several humps in the area that with only  a little
  digging  reveals more drums, forcing extensive excavation.
  There are also a  few abandoned truck trailers  containing
  both liquids and solids.
    In the typical case, there will also be multiple sources of
  potential contamination to nearby residents: a stream run-
  ning through the area feeding into a river near a city's water
  supply;  a children's summer camp; or signs of past fires.
  These raise fears of direct contact. When odors are present,
  they suggest the presence of volatiles with the  attendant air
  contamination.
  This description is only exaggerated to the extent that all the fea-
tures are thrown together. It is not unusual or atypical.
  Here is another situation:
    A ship going down a river has hit a bridge abutment which
  was supporting a train with a number of tank cars and box-
  cars all  holding various chemicals.  At least two of the cars
  have fallen  into the river below and broken open. Several of
  the cars still on the bridge and just beyond its confines have
   also broken open and are leaking.  There are what are de-
   scribed as fumes above the river, and a multi-colored plume is
  moving  down stream. A cloud is rising above the wreck site
  and there is a small fire beginning.
  These two situations, while vastly different in many ways, are
alike in that both must be evaluated and cleaned up, and the work-
ers must have a reasonable chance of doing that work without in-
jury or illness.
EMERGENCY MEDICAL ADVICE
  The time pressures of the emergency situation do not relieve the
physician of the necessity of giving advice that will: (1) protect the
workers, and (2) allow the cleanup work to proceed efficiently.
  However, there will never be enough  information; one always
wishes to know more. In most cases, one will not know all of the
substances present and even less about what mixtures may have
been formed  and what their toxic  effects might  be. Other  un-
knowns might include what water might do to the chemical's toxic-
ity or whether sunlight and variable climatic conditions  cause an
increase in toxicity.
  As they sit for years in drums,  do some  chemicals,  normally
stable, become unstable and  dangerous  if shocked or jarred? A
constant  danger  is that many compounds, the organophosphate
pesticides, for instance, are very efficient skin penetrators. So are
most of the solvent materials that are ubiquitous at hazard waste
sites.
  There are many useful databases available for these situations,
but not one ideal. Many of the questions are unanswerable because
no one knows what happens  under some of these circumstances.
Thus, one frequently must take what is  known or what can rap-
idly be obtained, make quick mental extrapolations and  then for-
mulate advice  and recommendations as to the personal protective
equipment and special procedures needed by workers at the site.
  Obviously, the protection of workers takes on a slightly different
character than is found in the usual industrial situation.

MEDICAL APPROACHES
  As with any medical issue, there are different ways of looking at
the situation.  Two extreme approaches will be  examined which I
will label the "Do-Everything" or "Do-Nothing" view.
Do-Every thing

  The Do-Everything approach is based on the assumption that
baseline tests should be carried out for as many substances as are
technically feasible. This approach also suggests baseline tests of as
many organ systems  as possible to  allow a comparison, if ex-
posure occurs, to determine the presence of harmful effects.
  There are problems with this approach.  Besides the  obvious,
enormous expense of such testing, there is the time that the worker
is off the job and the resulting loss of productivity. An employee
of the USEPA or one of the more active private contractors do-
ing this work may be involved in dozens of different sites each year,
each with its unique situations. It is difficult, if one believes in this
approach, to pick a frequency of testing that will allow the kind of
close watch that  is desired. Another major problem  with this ap-
proach is the very practical one of finding qualified and competent
laboratories and medical professionals within a reasonable distance
and able to do the testing required.
Do-Nothing

  The Do-Nothing view maintains that only those tests which will
show that a worker is in good general health should be carried out
                                                                                          PERSONNEL SAFETY
                                                        259

-------
and that continual emphasis on training and proper work  pro-
cedures be maintained. When an exposure occurs, the toxicity of
the substance or the situation is evaluated and appropriate, spe-
cific testing is carried out to determine "if any adverse effects  have
developed.
  With time, I  have come to lean toward the latter, or Do-Noth-
ing school of thought, although I would prefer that it be labelled
the "Do-A-Little-Bit, Well" approach. Here is my reasoning:
  First, the chances are significant that, even with an attempt to
measure everything, the substance to  which an exposure actually
does take place will not have been measured in the base-line tests.
Second, there is no chance whatever  that the combinations and
mixtures of substances can be directly evaluated. Third, there is the
tendency of multiple-testing programs  to give a false sense  of
security to both the workers and  the physicians responsible for
them. Finally, there is the very significant expense of the testing.
  A problem with the do-nothing  approach is that  little tangible
evidence is given to the worker or his  employer that his health is,
in fact, being monitored.
  What, then, is suggested as a medical surveillance program?

SURVEILLANCE PROGRAM
   A worker new to hazardous waste work should have a pre-place-
ment examination consisting of a complete and detailed occupa-
tional and medical history followed by a thorough physical exam-
ination made available to him. Emphasis during the physical should
be  placed on overall physical  fitness and suitability  for heavy
manual labor with a variety of tools and equipment and the ability
to wear and utilize a large variety of personal protective equipment
including SCBA and fully encapsulated suits. Vision  tests and pul-
monary function testing, including, as a minimum,  the FVC and
the FEV, are important.
   Although I feel that the utility is very limited, a postero-anterior
chest X-ray may be useful for comparison or baseline purposes.
   Since many of the substances encountered are hepato-toxic, a
multi-channel blood chemistry can be done, if available, since it
is often cheaper than individual tests, is widely reproducible and
contributes a nice range of information.
   If it is known that a  strong potential for exposure to organo-
phosphate pesticides exists, it may be worthwhile to perform a
baseline serum and/or red blood cell  cholinesterase. Likewise, if
other particular substances are known to be  present and the ex-
posure at that site will be of long duration, baseline tests specific
to them can be utilized.
  Other laboratory work-up is limited to a complete blood count
and urinalysis (with microscopic).
  The history should be updated yearly to determine if any ex-
posures may have had the potential for adverse effects. A brief
physical is also appropriate as is a  repeat of the pulmonary func-
tion test. If desired, the blood work can also be repeated. If a part
of the program, the chest X-ray can be  repeated perhaps every four
to five years.
  As you can see, my suggestions are quite variable and very site-
specific, but I believe that they will allow the responsible physician
to maintain a reasonably close view of the worker's  health with a
minimum in outlay of both funds and productivity of the employee.


MEDICAL EMERGENCIES

  For the emergency situation, there are several things that usually
demand consideration at the same time: the effects that may  be
occurring to those first responders (usually fire or police person-
nel); the workers brought in to perform the  cleanup;  effects to
those involved with the incident itself,  but close-by and thus  sub-
ject to its effects; and finally,  some peripheral issues concerning
"incidental" personnel such as reporters, etc., and  requests from
hospitals and local health departments for assistance.
  These situations are almost always chaotic,  particularly in the
early hours. Information is scarce, and the data that arrive are
                                                        sometimes contradictory. If some people on-site are overcome, this
                                                        situation creates more concern and the local medical care facilities
                                                        may be overwhelmed.
                                                          If one tries to wait for all of the desired information to arrive,
                                                        a long  time can elapse with  resultant unnecessary adverse health
                                                        effects  to the personnel involved. Therefore, as noted above, one
                                                        must act on what one knows, and make recommendations for pro-
                                                        tective equipment, work practices and evacuation on the best avail-
                                                        able data; however, one must also  be  prepared to  change those
                                                        recommendations, perhaps radically, when new data arrive.
                                                          When making treatment recommendations, there  are relatively
                                                        few chemicals(cyanidc or certain pesticides, for example) requiring
                                                        specific and/or immediate therapy. For most,  symptomatic  and
                                                        supportive care will serve the patient well until better information
                                                        becomes available. This advice makes emergency room personnel
                                                        more comfortable, since they tend to assume that every chemical
                                                        involved in an environmental incident requires specific and perhaps
                                                        exotic,  care. There is time, in most cases, for specialized care to be
                                                        instituted after the emergency has calmed down a bit.
                                                          It should be mandatory that patients treated and released be told
                                                        to keep close awareness of their physical condition, particularly the
                                                        pulmonary system, for a period of 72 hours and to report to med-
                                                        ical facilities if changes are noted.
                                                        Contrast with Industry
                                                          In contrast to the usual industrial situation where workers use
                                                        known  specialized equipment, techniques or substances and have
                                                        special  testing or examinations specifically tailored to their poten-
                                                        tial exposures or duties made available, the working environment
                                                        of the hazardous waste worker is more difficult to define. In indus-
                                                        try one can, if necessary, work backward and discover those sub-
                                                        stances or conditions to which the workers may have been exposed.
                                                        There are always many  things one  does  not know,  but there is
                                                        usually a place to begin. In the hazardous waste situation, this may
                                                        not be true.
                                                          When wearing the fully encapsulated  "moon suit" and utilizing
                                                        SCBA, the worker is also clumsy and unable to move quickly, thus
                                                        creating a potential  for  trauma-producing accidents. It is widely
                                                        accepted that far more injuries on hazardous waste sites are caused
                                                        by accidents than by chemical exposures.

                                                        Weather
                                                          Since it is  obvious that most cleanup takes place outdoors, the
                                                        weather plays  a large role. The toxicities of the substances  can
                                                        vary; cold weather, for example,  reduces vapor pressure and there
                                                        is less tendency for substances to volatilize; heat causes the opposite
                                                        effects.
                                                          Extreme heat or cold creates  worker stress,  as in any out-of-
                                                        doors employment,  but temperature is an  extra problem  to the
                                                        hazardous waste worker primarily because of the extensive protec-
                                                        tive equipment that he  must wear. In summer, significant heat
                                                        loads are generated quickly because of the impervious clothing
                                                        utilized. The workers on the site and  their supervisors must be
                                                        familiar with the signs and symptoms of heat stress and know the
                                                        appropriate first aid procedures.
                                                        CONCLUSIONS
                                                          In summary, the hazardous waste program in the United States
                                                        is one that presents many fascinating challenges to the occupational
                                                        physician. Meeting those challenges requires an ability to innovate
                                                        and make decisions quickly if the workers are to have the protec-
                                                        tion to  which they are entitled.

                                                        DISCLAIMER
                                                          The  opinions expressed in  this paper do not represent official
                                                        NIOSH or ATSDR policy and should  in no way be taken as an
                                                        official position of the U.S. Government. They are  the author's
                                                        viewpoint developed through experience with these situations; one
                                                        that he has found to be useful.
260
PERSONNEL SAFETY

-------
     MULTIPLE  CHEMICAL  EXPOSURE  CONSIDERATIONS
 ASSOCIATED  WITH HEALTH  AND  SAFETY  ASSESSMENTS

                                     THOMAS C. MARSHALL, Ph.D.
                                                 IT Corporation
                                             Knoxville, Tennessee
INTRODUCTION

  Multiple chemical exposures exist in many industrial and en-
vironmental situations, but the problems associated with enhanced
toxicity (toxic interactions) due to combined exposure are often ig-
nored. Hazardous waste sites present a worst case potential in this
regard due to the multitude of different chemicals often found at a
single site. The possible combinations are enormous and largely un-
predictable. In this paper, the author discusses a set of principles
governing toxic interactions which can be used to recognize and
evaluate the potential hazards of waste sites by those responsible
for employee health and safety and by those conducting health risk
assessments. Although interactions between two chemicals can lead
to decreased toxicity, only the concern of increased toxicity is
discussed.
  Toxic interactions are manifested by an effect on one or more of
the major physiologic processes that determine how the  body
handles foreign materials.' These processes are the absorption,
distribution, biotransformation and  excretion of chemicals. This
section is followed by a discussion of recommendations on how to
manage the risk of toxic interactions once the potential hazard of a
multiple chemical exposure is known.

BASIC CONCEPTS
  Simplified, toxic interactions  can arise by the two  general
mechanisms of addition and potentiation. Additive interactions are
more easily conceptualized. For example, many chemicals such as
chloroform and diethylether cause central nervous  system depres-
sion manifested by the typical inebriating effects of uncoordination
and light-headedness. A specified exposure to any one of  these
agents may cause slight or no detectable adverse effects. However,
if one was exposed to several of these agents simultaneously, diz-
ziness or some other effect could ensue even though the exposure
concentrations of the individual chemical components were the
same as when the person was exposed to that chemical alone with
little or no effect. This type of interaction is usually additive. That
is, the combined effect is equivalent to the sum of the effect when
each agent is given alone. In the exceptional circumstance when the
combined effect is much greater than that predicted from the sum
of the two, this is referred to as synergism. The mechanisms behind
many synergistic interactions are unknown.
  The toxic interactions termed potentiation are situations where
one substance does not cause a certain toxic effect, but when com-
bined with another substance that does cause that effect, the toxici-
ty is greatly enhanced. For example, isopropyl alcohol is not toxic
to the liver while carbon tetrachloride is hepatotoxic. When  these
                         Figure 1
 Schematic representation of the body showing how the different organ
 systems interact in the processes of absorption, distribution, biotrans-
 formation and excretion.
two materials are combined, the hepatotoxic response to exposure
is much greater than when carbon tetrachloride is given alone.2
This interaction has been documented in man as a result of an in-
dustrial exposure to both agents.3
  A brief overview  of the dynamic processes involved in how
foreign materials are absorbed, handled and excreted will help pro-
vide a background for understanding how toxic interactions of
potentiation occur. Figure 1 contains a schematic representation of
the body and some of its components. Exposure to a chemical can
occur by various routes such as  skin contact, ingestion and inhal-
ation.
  In these cases, the layers of skin, lining of the gastrointestinal
tract and membranes, in the  lung serve as barriers between the
chemical and the general circulation. Crossing these barriers is
referred to as absorption.  Once a chemical gains access  to the
blood, it is distributed to organs throughout the body by the cir-
culatory system. As shown in  Figure  1, some  organs actively
eliminate chemicals from the body: the kidneys via urinary excre-
tion, the digestive system via fecal excretion and the lungs via
gaseous exhalation.
                                                                                             SITE SAFETY
                                                     261

-------
  If a foreign compound is not readily excreted in its parent form,
the body is capable of transforming the chemical, usually rendering
it more water soluble, so that it is more easily excreted. This process
of biotransformation is a major function of the liver.
  These processes of absorption,  distribution, biotransformation
and excretion can be affected by the toxic actions of one chemical
exposure in  a manner making the body more  susceptible to other
chemicals during the same time period of exposure or in a subse-
quent exposure.
  This is  the basic concept that  accounts for most instances of
potentiation and explains some examples of synergism. Most of the
published literature addressing toxic interactions involve studies on
pharmaceutical agents,  while little  specific information in  this
regard exists on industrial chemicals. As much as possible, this
paper describes the few studies describing toxic interactions among
industrial chemicals relevant to  the hazardous waste industry.

RECOGNITION  OF POTENTIAL
TOXIC INTERACTIONS
Additive Interactions
   Potential  additive interactions  are easier to recognize  because
they usually involve substances that share the same target organs
for toxicity. All organ systems are  potential targets of toxic interac-
tions caused by multiple chemical exposures.  Particularly suscep-
tible, though,  are the respiratory tract as the major route of oc-
cupational exposure, the  liver as  the principal  organ in the
biotransformation of foreign compounds and the kidneys as the
predominant route of foreign compound excretion.
   Candidates for additive interactions involving acute lung injury
and chronic  pulmonary  fibrosis are: acid  fumes and other
pulmonary irritants such as the metal fumes of cadmium, beryllium
and aluminum; fibrogenic dusts such as asbestos and silica; and
oxygen in abnormal concentrations.4 The liver is sensitive to many
chlorinated  hydrocarbons  including  polychlorinated  biphenyls
(PCBs) and  chlorinated  dibenzofurans (structural analogs  to diox-
ins), which are extremely toxic  and  are commonly found as con-
taminants of  PCBs.  The  kidneys  are also  sensitive to several
chlorinated  hydrocarbons (i.e.,  chloroform  and  carbon
tetrachJoride) and heavy metals (i.e., mercury and chromium).

Potentiation Interactions

Absorption as a site.
  Occupational poisoning cannot occur without absorption of a
toxicant. Any  chemical  exposure affecting the barriers to absorp-
tion by making them more permeable is a candidate for poten-
tiating the toxicity of other chemical exposures by enhancing the
quantity of those chemicals absorbed. For example, the skin, due
to its thickness and structure, poses a significant barrier to penetra-
tion by  chemicals,   especially  those that  are  water  soluble
(hydrophilic). However, a damaged skin is more readily penetrated
by  both  hydrophilic  and  lipophilic  (fat  soluble) substances.
Therefore, washing the hands with organic solvents (i.e., thinners
and gasoline) or with abrasives which damage the skin increases the
risk of skin penetration by other toxic substances. Similarly, surfac-
tants used as foaming, dispersing,  wetting, detergent and emulsify-
ing agents and phenols disintegrate cellular membranes in  the skin
and decrease the skin's  ability to limit the absorption of foreign
compounds.
  The skin is an important route of occupational exposure, but the
respiratory tract is the major pathway. The  lung  is designed to
facilitate rapid absorption  due to  its function  of gas exchange, so
there is little that can happen  which would  further enhance its
permeability. Gardiner  and Schanker5'6'7'8  showed  that  rats in-
haling acidic fumes, paraquat, aerosolized palain and rock dust all
suffered irritancy  and damage which  appeared  to increase the
porosity of the pulmonary epithelium and enhancing absorption.
Also, ciliated cells in the respiratory airways help keep the lungs
clear of foreign particles. There are agents, such as formaldehyde
and cigarette smoke, which  strongly inhibit this ciliary action of
                                                        bronchial  epithelium  which  reduces  the efficiency of  particle
                                                        removal. Consequently, there is an increase in the quantity of par-
                                                        ticles retained in the lung and a potential increase in the amount of
                                                        foreign material absorbed from within or on those particles.

                                                        Biotransformation as a site.
                                                          This is probably the most common site for toxic interactions due
                                                        to the phenomena of enzyme inhibition and enzyme induction.
                                                        Literally all foreign compounds are subject to biotransformation
                                                        by enzyme systems which are  primarily  located in the liver but are
                                                        present to some extent in all organs of the body. Biotransformation
                                                        reactions usually detoxify foreign compounds and make them more
                                                        readily excreted.*  However, it is common for these reactions  to
                                                        form  toxic intermediate or end  products (metabolites) that are
                                                        more  toxic than the parent materials comprising the original ex-
                                                        posure.  Therefore, any alteration in the enzyme systems respons-
                                                        ible for biotransformation, either inhibition or  induction, can
                                                        cause  a toxic interaction depending on the nature of the materials
                                                        involved in the multiple chemical exposure.
                                                          Unfortunately, most of the  studies that have been conducted on
                                                        toxic interactions involving biotransformation have been designed
                                                        to elucidate mechanisms and have not been conducted by routes of
                                                        exposure or at dose levels that resemble "real life" circumstances.
                                                        Also,  therapeutic  agents  and not industrial chemicals are those
                                                        most frequently studied. A few illustrative examples pertinent to in-
                                                        dustrial chemicals are given in this paper and the interested reader is
                                                        referred  to a couple of exhaustive reviews.10'"
                                                          Enzyme inhibition leads  to toxic interactions by slowing  the rate
                                                        at which a toxic foreign compound is metabolized, thereby  increas-
                                                        ing the systemic exposure to the parent  material. Since metabolism
                                                        of a chemical facilitates its excretion from the body, biotransfor-
                                                        mation inhibitors also increase the residence time of other foreign
                                                        substances in the body.
                                                          Two chemicals may inhibit one or the other's biotransformation
                                                        by competing for the same limited enzyme.  This  is true of short-
                                                        chain  alcohols  and glycols  which are all  metabolized by  alcohol
                                                        dehydrogenase. Some of the best inhibitors  of biotransformation
                                                        enzymes are those that undergo little or no metabolism, such as the
                                                        perfluorinated hydrocarbons.  Other compounds are commercially
                                                        important specifically because they inhibit metabolism and, thus,
                                                        augment the  activity of insecticides. The 1,3-benzodioxles are the
                                                        best examples;  butoxide is the most commonly used.'1 Other im-
                                                        portant inhibitors  are  sulfur-containing compounds, such as car-
                                                        bon disulfide and the phosphorothionates" and some nitrogen-
                                                        containing compounds, such as the imidazoles."
                                                          Enzyme induction can cause toxic interactions by increasing the
                                                        rate at which a non-toxic foreign compound is converted into a tox-
                                                        ic form.  Many of the most effective enzyme  inducers are aromatic
                                                        chlorinated hydrocarbons with prolonged biological half-lives (i.e.,
                                                        slowly excreted from the body). These include the chlorinated in-
                                                        secticides and PCBs. The  chlorinated dibenzodioxins and diben-
                                                        zofurans are  the  most  potent  inducers  known.  Other well
                                                        documented inducers include the polynuclear aromatic carcinogens
                                                        and certain drugs,  most notably phenobarbital.
                                                          Another important feature of biotransformation enzyme induc-
                                                        tion is its onset and duration. The inducing action of different com-
                                                        pounds begins at different rates and continues after cessation of ex-
                                                        posure for various lengths of  time. The time after exposure when
                                                        maximum induction occurs varies from  one day for 3-methylchol-
                                                        anthrene to approximately 2 weeks for the insecticide chlordane."
                                                        The duration of induced  enzyme activity can be  from days to
                                                        months after the last exposure.

                                                        Elimination as a site.
                                                          Renal elimination  is the major  route of excretion  for many
                                                        foreign  compounds.  Any  intoxication  which interferes with this
                                                        process has the potential for causing a toxic interaction by allowing
                                                        the accumulation of chemicals in the body during a combined ex-
                                                        posure or during an exposure subsequent to the renal injury. The
                                                        glommerular membrane of the kidney is extremely porous and
262
SITE SAFETY

-------
allows all but the high molecular weight materials, such as proteins,
to pass through. The subsequent tubular epithelium is a lipoprotein
barrier where water and specific materials, for example electrolytes,
are reabsorbed. The metabolities of foreign compounds are usually
not reabsorbed but pass on into the urine unless the tubular mem-
brane is damaged. Nephrotoxic agents which damage the proximal
tubular epithelium include short  chain chlorinated hydrocarbons
(chloroform, carbon tetrachloride) and certain heavy metals (mer-
cury, chromium, uranium).

MITIGATING FACTORS IN TOXIC INTERACTIONS
  Several factors can make toxic interactions more likely to occur.
These include preexisting disease states of the liver,  kidneys and
lungs which would make a person more susceptible to the many
toxic effects that lead to toxic interactions. Prescribed and abused
drugs, such as alcohol, are difficult to control and monitor, yet
these have a high potential for being involved in an interaction.
Whereas occupational exposures are carefully controlled to prevent
any detectable toxic  effects, drugs are  taken deliberately in quan-
tities high  enough  to elicit an altering  effect.  For  example,
phenobarbital  at therapeutic  doses  is  a potent  inducer  of
biotransforming enzymes, an effect that may cause a toxic interac-
tion. Also, Elovaara et a/." showed that inhalation of xylene when
coupled with ethanol ingestion produced severe liver damage, while
independent exposure to xylene or ethanol failed to  do so.
 MANAGEMENT OF RISKS TO INTERACTIONS
   Toxicity due to  occupational exposure is generally manifested
 only during episodes of abnormally high exposure concentrations.
 Most industrial hygiene programs limit occupational exposures to a
 very significant extent, thus greatly reducing the potential for toxic
 interactions.  Their  consideration   is  nonetheless  important,
 especially in circumstances where operations are being conducted at
 or near the occupational  exposure  limits,  and/or the multiple
 chemical exposure will occur  over an extended period of time.
 Several recommendations given below should help control the risks
 to toxic interactions resulting  from multiple chemical exposures.
 Depending on the sophistication of a company's industrial hygiene
 program, these may be incorporated into existing activities without
 any additional costs.
 Evaluation
   Each situation involving  multiple chemical exposures should be
 evaluated for the potential of a toxic interaction in view of the prin-
 ciples  discussed  in this  paper. Depending on  the  chemicals  in-
 volved,  this  will require  seeking out  additional information and
 referring to more sources than would otherwise be necessary. The
 chemicals should be categorized by  their routes and projected
 magnitudes of exposure, target organs for toxicity and potential to
 alter the biotransforming enzyme systems.  Input should  be ob-
 tained from a toxicologist,  if possible.
   The  time required for  this  qualitative assessment is more
 justifiable on a  cost-effective  basis for circumstances where the
 multiple chemical exposure is routine and the same substances are
 consistently involved. For situations where the exposures are con-
 stantly changing, the evaluation is an on-going process and more
 time consuming but just  as important as in the case of the more
 routine work conditions. If such evaluations are not possible, then
 a very conservative approach should  be taken toward protecting
 workers against exposure.
   The American Conference of Governmental Industrial Hygien-
 ists" has developed a threshold limit criterion for chemical mix-
 tures and multiple  chemical exposures. The quantitative approach
 described is adequate for circumstances where there is no interac-
 tion or where the interaction is additive. However, the approach
 may not be adequate to control potentiation interactions, and the
 mathematical manipulations should  be  adjusted accordingly de-
 pending on the details of the particular situation.
  The NAS (1980) proposed a quantitative method for compen-
sating for known instances of toxic interactions. However, this re-
quires sufficient health effects data to quantitate the potentiation,
and only in exceptional circumstances will such data be available.
In most cases, adjustments in the limits to exposure deriving from
concerns over possible toxic interactions can be made only through
a subjective evaluation.

Medical History and Monitoring
  Many  companies in the hazardous  materials industry already
have occupational medicine programs that include obtaining de-
tailed medical  histories  and obtaining periodic medical exams.
These records should be consulted when individuals are assigned to
different work  projects with the potential for chemical exposure,
and the accuracy of the information should be checked with par-
ticular attention given to prescription medications and chronic ill-
nesses.  Incorporation of medical  monitoring  into  work  plans
should be considered when the proposed work activities will involve
extremely toxic materials.

Training
  Education on the importance of an accurate medical history, in-
cluding the use of prescription medications, should be incorporated
into  the company's  health  and safety  training  program.  The
necessity  of  occasional medical  monitoring  for  certain  work
assignments should be explained to employees. This step will aid in
obtaining employee cooperation and will  decrease the chances of
alarm  when  a  situation requiring  monitoring is encountered.
Employees should be made aware that the exposure scenario will be
re-evaluated when new  chemicals are introduced into their work
activities.

Workplace Monitoring

  Sufficient work area monitoring of  chemicals should be  con-
ducted to  ensure that  exposures  to multiple  chemicals are  not
higher than anticipated. The results from monitoring operations
should be reviewed frequently to detect changes. If new chemical
substances are introduced into the occupational environment, the
situation should  be reevaluated with consideration given  to the
potential for  toxic interactions.
REFERENCES

 1.  National Academy of Sciences, Principles of Toxicological Interac-
    tions Associated with Multiple Chemical Exposure, NAS Press, Wash-
    ington, D.C., 1980.
 2.  Plaa, G.L., Traiger,  G.J., Hanasono, O.K. and Witschi, H.P., "Ef-
    fect of alcohols on various forms of chemically induced liver injury;"
    In Khanna, J.M., Israel, Y., and Kalant, H. (eds.): Alcoholic Liver
    Pathology. Addiction Research Foundation, Toronto, Ontario, 1975,
    255-44.
 3.  Folland,  D.S., Schaffner, W., Grinn, H.E., Crofford, O.B.  and
    McMurray, D.R., "Carbon tetrachloride toxicity potentiated by iso-
    propyl alcohol, J.A.M.A., 236,  1976, 1853-56.
 4.  Morgan, W.K.C., and  Seaton, A., Occupational Lung Diseases.
    Saunders, Philadelphia, PA, 1975.
 5.  Gardiner, T.H. and  Schanker, L.S., "Effect of papain-induced em-
    physema on permability of rat lungs to drugs."  Proc. Soc.  Exp
    Biol. Med. 149, 1975, 972-977.
 6.  Gardiner, T.H. and Schanker, L.S.,  "Effect  of  paraquat-induced
    lung damage on permability of rat lung to drugs," Proc Soc  Exp
    Biol. Med., 151, 1975, 288-292.
 7.  Gardiner, T.H. and  Schanker, L.S., "Effect of oxygen toxicity and
    nitric acid-induced lung disease on drug absorption from the rat
    lung," Res. Commun. Chem. Pathol. Pharmacol., 15, 1976, 107-120.
 8.  Gardiner, T.H. and  Schanker, L.S., "Enhanced pulmonary absorp-
    tion of drugs in rats with experimental silicosis." Res. Commun
    Chem. Pathol. Pharmacol., 13,  1976, 559-562.
                                                                                                        SITE SAFETY
                                                                                                                             263

-------
 9. La Du,  B., Mandel,  H.C. and Way, E.L.,  Fundamentals of Drug
    Metabolism and Drug Disposition. Williams and Wilkins, Baltimore,
    1971.
10. Testa, B. and Jenner, P., "Induction and inhibition of drug-metabol-
    izing enzyme systems." B. Testa and F. Jenner, eds. Drug Metabol-
    ism: Chemical and Biochemical Aspects. Marcel Dekker, New York,
    NY, 1976, 329-350.
11. Bock, K.W. and Remmer, H., "Introduction to  hepatic hemopro-
    teins." F. DeMattels and W.N. Aldridge, eds. Handbook of Experi-
    mental Pharmacology, 44. Heme and Hemoproteins. Springer-Verlag,
    Berlin, 1978, 49-80.
12. Philpot,  R.M.  and  Hodgson, E., "A cytochrome P-450-piperonyl
    butoxide spectrum similar to that produced by ethyl isocyanide,"
    Life Sci.  10, Pt. II, 1971, 503-512.
13. Neal, R.A., Kamataki, T., Hunter, A.L. and Catignani, O., "Monox-
    ygenase catalyzed  activation of thiono-sulfur containing compounds
                                                                to reactive intermediates. V. Ullrich, A.  Hildebrandt, I. Roots, R.W.
                                                                Estabrook, and A.H. Conney,  eds. Mlcrosomes and Drug Oxida-
                                                                tions. Pergamon Press, New York, NY,  1977, 467-475.
                                                             14. Wilkinson, C.F., Hetnarski, K., Cantwell, O.P. and Di Carlo, F.J.,
                                                                "Structure-activity relationships in the effects of 1-alkylimidazoles on
                                                                microsomal oxidation in vitro and  in vivo." Biochem. Pharmacol.,
                                                                23, 1974. 2377-2386.
                                                             15. Elovaara,  E.. Collan, Y.,  Pfaffli, P., and Vainio, H.. "The  com-
                                                                bined toxicity of technical grade xylene and ethanol in the rat," Xeno-
                                                                biotica 10, 1980, 435-445.
                                                             16. ~, TLVs: Threshold Limit Values for Chemical Substances and Physi-
                                                                cal Agents in the Workroom Environment with Intended Changes for
                                                                1983-84,  Proc.  American  Conference  of  Governmental  Industrial
                                                                Hygienists, Cincinnati, OH 1983.
264
SITE SAFETY

-------
              CONTROL  OF FUGITIVE DUST EMISSIONS  AT
                     HAZARDOUS WASTE  CLEANUP  SITES

                                              KEITH D. ROSBURY
                                                PEI Associates, Inc.
                                                 Golden, Colorado
                                              STEPHEN C. JAMES
                                Municipal Environmental Research Laboratory
                                    U.S. Environmental Protection Agency
                                                 Cincinnati, Ohio
 INTRODUCTION

   Spills, waste disposal and various industrial operations can re-
 sult in the contamination of land surfaces by toxic chemicals. Soil
 particles from these areas can be entrained into the airy transported
 off-site by the wind and result in human exposure by direct inhala-
 tion. Indirect exposure could result if particulates are deposited in
 agricultural fields, pastures or waterways and thus enter the human
 food chain. Since many environmentally troublesome compounds
 are tightly bound to particles, and many surface-contaminated sites
 have conditions favoring wind  erosion, such as  sparse vegetation
 cover and high levels of activity which disturb the surface, this ex-
 posure route is important.
   Contaminated soil can be entrained by the air in three basic
 ways:
 •Wind erosion
 •Reentrainment by moving vehicles (rubber-tired or tracked  ve-
 hicles) on soil or paved/unpaved roads
 •Active cleanup (movement of soil by dozers, loading by front-
 end loaders, etc.)
   These three mechanisms can act individually or in combination.
   Dust control at a hazardous waste site is a different problem than
 control of non-contaminated dust to improve particulate air qual-
 ity. While 50 or 75% control of dust from an unpaved road might
 be adequate for air quality purposes,  it is not adequate for con-
 taminated  dust. Any direct or indirect human exposure to con-
 taminated  soil is  potentially harmful; 100% control is the  de-
 sired goal.
   Very little information is available to assist in developing dust
 control programs at cleanup sites.  Field  testing of  dust control
 effectiveness has been limited to vehicle-caused re-entrainment of
 dust from paved and unpaved roads. The purpose of this project
 was to:
 •Perform field demonstrations of several products to determine
 their effectiveness for controlling dust
 •Prepare a handbook on state-of-the-art methods  of dust control
  Three field studies were performed. In these studies, 14 dust sup-
 pressants were tested to determine their effectiveness in  controll-
 ing fugitive dust against  wind  erosion from exposed areas. The
 second wind erosion field study was an evaluation of the effective-
 ness of windscreens, and windscreen/dust suppressant combina-
 tions, in controlling  fugitive dust  from storage  piles. The third
 field study, investigating active cleanup emissions, consisted of test-
ing fugitive dust control  measures applicable to loading dirt by
front-end loader into a truck.
  A Dust Control Handbook was prepared. For each of the three
basic re-entrainment mechanisms, the following were described:
•Identification of dust producing points
•Principles of control
•Product listing by name, address, telephone  number, dilution,
 application rate, basic application method and cost
•Detailed application procedures
•Product effectiveness

DATE COMPILED
Control of Wind Erosion Dust Emissions
from Exposed Areas
  The objective of this study  was to identify dust suppressant
methods that are 100% effective in controlling wind erosion emis-
sions. This  criterion allowed flexibility in designing the sampling
protocol since no control efficiency or emission rate data were re-
quired. The answer could be, "yes," the dust suppressant is 100%
effective, or "no," it is not 100% effective.
  Although several approaches were available for testing wind ero-
sion emissions, tracer sampling was  chosen as the methodology
most closely paralleling the requirements of this study. In order to
determine the effectiveness of a dust suppressant in controlling fug-
itive emissions from an exposed test plot, it is necessary to detect
particles leaving the plot. This may be done by capturing airborne
particles. To overcome the problem of determining the origin of
the loose material, a tracer was added to the soil of the test plot.
The tracer was mixed  with the soil before a commercial dust sup-
pressant product was applied to the surface. Any tracer-laden par-
ticles later found in the ambient air around the test plot indicated a
failure in the integrity of the crust formed by the dust suppressant.
  The test plots  were located on  a small farm near Cincinnati,
Ohio. The plots were  located several hundred feet apart to elim-
inate cross contamination of tracer between plots. The  plots were
prepared by removing vegetation from 50 ft x 50 ft areas with a
bulldozer and grading the plot smoothly. One of two different
tracers was applied to  each bare plot. Zinc oxide (water insoluble)
was applied at a rate of 0.04 lb/yd2. Zinc sulfate (water soluble)
was applied  at a rate of 0.01 lb/yd2. After application of the tracer,
the dust control products were applied according to the manufac-
turer's recommendations. Test plot data are given in Table 1.
  Particulates being removed from the plot by wind  movement
were sampled with saltation catchers, 36 in. plastic tubes with a
2 in. wide vertical slot sampling the 12 to 30 in.  height interval.
Four catchers were placed at each plot, oriented  at the midpoint
                                                                                                SITE SAFETY
                                                      265

-------
                                                              Table 1
                                                       Exposed Area Teal Plot*

Test
Plot
1
2
3
4
5
6
7
8
9

10



Plot Sue
50 ft. X 50 ft.
50 ft. X 50 ft.
50 ft. X 50 ft.
50 ft. X 50 ft.
50 ft. X 50 ft.
50 ft. X 50 ft.
50 ft. X 50 ft.
50 ft. X 50 ft.
50 ft. X 50 ft.

50 ft. X 50 ft.



Tracer
zinc sulfate
zinc sulfate
zinc sulfate
zinc sulfate
zinc sulfate
zinc sulfate
zinc sulfate
zinc sulfate
zinc oxide

ztnc oxide


Pre-
emergent
no
no
no
no
no
no
no
no
no

no

Dust Suppressant Tested

Name
Soil Seal
AMSCO-RES 4281
Fiber mat
F Iambi nder
Genaqua
Curasol
H166 & M167
Coherex CRF
Sherman Process (mulch,
no grass seed)
Sherman Process (mulch.
with grass seed)
Application
Concentration
0.03:
20t
8 oz./yd*
1 71
101
31
61
25»
..

..

Application
Date
1.0 gal/yd'
0.6 gal/yd'
3 12- foot rolls
0.5 gal/yd*
0.2 gal/yd*
0.3 gal/yd'
0.5 gal/yd*
0.5 gal/yd»
--

--

                                                             Table 2
                                                   Saltation Sampler Rcnulu (ppm)

Test
Plot
1
2
3
4
5
6
7
8
9
10

Date
Established
5-24-84
5/11/84
5/18/84
5/25/84
5/18/84
5/?4/84
5/24/84
5/24/84
7/10/84
7/10/84
Sample Date

6-11
78
71
42
69
91
75
64
113



6-25
97
114
163
105
172
111
76
162



7-3
..
44
74
35
176
42
42
36



7-13
111
121
125
116
152
130
96
100



7-20
85
121
65
55
90
67
67
72



7-25
311
170
<30
—
65
58
151
454



7-30
155
322
<65
166
58
97
152
118



8-8
101
111
28
219
128
97
156
174



8-22
183
172
75?
140
141
167
160
206
513
157

8-31
170
204
120
119
217
128
122
214
132
103
  of each side with the sampling slot facing the plot. Samples from
  the saltation catchers were taken roughly once each week during
  the test period and were analyzed for zinc by atomic absorption
  spectroscopy. The  results  of the saltation sample analyses  are
  shown in Table 2. Values over 75 ppm  (the background concen-
  tration of zinc in the soil tested in each plot) indicate failure of the
  dust suppressant.
   The data indicate integrity of the crust on all plots on June 11
 except Plots 5 and  8. Time since application varied from 17 to
 31 days. Two weeks later, on June 25, only Plot 7 remained near
 the background  level of zinc in the soil. Values remained above
 background  levels on most plots throughout the  period.  A no-
 table exception was the fiber mat where values decreased below
 background by July 20 and remained there for a month.
   A  problem with all plots was  the rapid  regrowth  of vegeta-
 tion.  All plots had  been stripped of vegetation before applica-
 tion of the tracer and dust suppressant. The rationale for this ac-
 tion was that a dust suppressant spray could not form a dirt crust
 in the presence of vegetation stems, because the dust suppressant
 would not uniformly pass the vegetation and reach  the soil. When
 vegetation did grow, it penetrated the crust, and a small dirt pile
 was seen around each stem. Testing of these dust  piles indicated
 large  quantities of the tracer material (> 200 ppm). Applying this
 finding to hazardous waste sites, it is apparent that  vegetation
 must be controlled if the crust is to stay intact.
  The alternate approach of promoting vegetation  as a dust con-
 trol measure is discussed in the  interpretation  of findings sec-
 tion.


Control of Wind Erosion Dust Emissions
from Storage Piles

  The field study effort was designed to demonstrate  the effec-
tiveness of a windscreen, alone or in combination with other con-
 trol measures, in reducing dust emissions due to wind  erosion
 from an inactive waste storage pile. Concurrent  upwind/down-
 wind aerosol/windspced measurements  were made with real-time
 data retrieval around a storage pile protected by a windscreen.
   The  waste  storage pile  tested  was  not  contaminated.  The
 assumption was made that  the  hazardous material would be in-
 tegrally bound to the soil and  would therefore be  emitted at a
 rate directly proportional to the loss rate of soil  from the pile.
 The soil material in the pile was relatively erodible so that detec-
 table concentrations  could  be obtained  during short-term tests
 even  with  moderate  (10  to 15 miles/hr) windspeeds. The  soil
 selected was a very fine shredded topsoil. The size  of the pile was
 established by determining a size large enough  to accurately sim-
 ulate wind erosion  action from a temporary waste  storage pile,
 yet not so large that it consumed a major portion of the study's
 resources. This  "reasonable" size  was estimated  at about 8 ft
 high with an elliptical base  25  ft by 20 ft, or about 100 tons of
 shredded topsoil.
  All atmospheric paniculate concentrations were  measured with
 RAM-1 continuous aerosol monitors. These instruments emit a
 pulsed light across  a continuous flowing sample  airstream and
sense the amount of light scattering with a silicon detector. Run-
ning side by side, the five RAM-Is  employed in this study had a
sampling precision of about 5 jig/m' when measuring ambient
 levels. The Ram-Is are stated to have an upper  particle size range
 of 20-jLim diameter, but are generally considered to  have a 50%
 collection efficiency upper cut point of about lOjim.
  Windspeed and direction  were measured with a Met One sys-
tem that included six windspeed sensors and one direction vane.
The anemometer sensors had an accuracy of ± 5 degrees.
  Both the  RAM-Is and  wind instruments   were  connected
through  a  translator  and an analog-to-digital convertor  to  an
Apple He computer onsite  that collected  the  individual  signals
during testing,  displayed  the  readings  continuously,  averaged
266
         SITE SAFETY

-------
INTERPRETATION OF FINDINGS

Control of Wind Erosion Dust Emissions
from Exposed Areas

  The  dust  suppressants applied cost  between $62 and $48407
acre in material cost, with a median cost of less than $800/acre.
Based on tracer studies, 100% effectiveness varied from less than
17 days to 30 days.  No re-applications were tested. However, it
is reasonable to assume that control after re-application would be
slightly greater than with the initial application.
  The  problem of weed control must be considered. Weeds punc-
tured the crust formed by the dust suppressant. Around each stem
was a small dirt pile which was highly contaminated with tracer.
Spraying preemergent  weed control on the plot before applica-
tion of the dust suppressant largely solved this problem.
  An alternate procedure to eliminating vegetative growth would
be to encourage it. Products are available that are temporary soil
binders impregnated with grass seed. While grass was beginning
to  grow,  the same  weed problem  just described would  occur.
Assuming a thick stand of grass did grow, control would prob-
ably not be  100% since there would always be some loose dirt be-
tween  grass stems. Chemical  dust suppressants sprayed on thick
grass stands would not be effective because the suppressant would
stick to the stems and little would penetrate to the soil.
 Control of Wind Erosion Dust Emissions
 from Storage Piles
   Test results show that the windscreen showed no consistent pos-
 itive control effectiveness  for particles in the <10 pm respirable
 range. Control effectiveness for particles  >10 um may be greater,
 but the study instrumentation could not  measure this size range.
 Particles in the  size range of 10-30 um  can stay suspended for
 distances of several miles. Particles  > 30 pm usually fall back to
 the ground within a few hundred feet of the source.
  Application of chemical dust suppressants to the pile were shown
to be about 50% effective within three days of suppressant appli-
cation. This effectiveness would decline  with time, but the decay
function was not established. Adding a windscreen upwind of a pile
treated with a dust suppressant showed no incremental increase in
control efficiency.
  If a windscreen is erected to control the 10-30 >im size fraction
(control efficiency unknown), the screen should be higher by 2 to 4
ft than the top of the pile, or negative control efficiencies may re-
sult at certain downwind distances.
  The combination of a windscreen and chemical dust suppres-
sant was unable to achieve  100% control. Therefore, some down-
wind exposure to contaminated dirt from storage piles can be an-
ticipated.
Control of Dust Emissions
from Active Site Cleanup
  Water spraying  over the area  being  worked by a front-end
loader and truck at the rate of 0.9 gal/yd2 resulted in a control
efficiency of about 50% for the front-end loader working area and
about 65% for the material dump area.  Adding surfactant to the
water allowed the use of less water (0.75 gal/yd2), while increasing
control  efficiency to about 70% for both the PEL working  area
and  PEL dumping area. Dryer conditions  than those experienced
at the test plot would require more water. It is unlikely that any
acceptable level of area watering would significantly increase  con-
trol levels measured. Therefore, the goal of 100% control efficiency
does not appear possible with this technology, potentially causing
subsequent human exposure impacts.
DISCLAIMER
  Although the research described in this article has been funded
in part by the USEPA through contract No. 68-02-3512, it has not
been subjected to the Agency's required peer and administrative
review and therefore does not necessarily reflect the views of the
Agency and no official endorsement should be inferred.
                                                                                                      SITE SAFETY
                                                                                                                          267

-------
them  over  1 min,  5  min,  and 1 hr time periods, and stored the
averages for later  retrieval. All the calibration  values could  be
changed by inputting new values to the computer.
  Both the upwind sampling stand and the main  downwind stand
had RAM-Is and windspeed sensors mounted at  3.3 ft and at 6.6
ft  heights.  The upwind stand also had the wind direction  vane
mounted at a  height of 6.6 ft. The third sampling  stand, which
was alternated between a location beside the  main downwind
stand and a further distance downwind, had a RAM-1 and wind-
speed sensor only at a height of 3.3 ft. A sixth windspeed sensor
was located at the upwind edge of the pile at a height of 3.3 ft.
  Results from the tests showed that the windscreen did not con-
sistently attain a significant level of control of particles Ł 10;im
(the inhalable  size  fraction measured by the RAM-1). The screen
was effective in reducing  windspeeds,  but this did  not result in
commensurate  reductions  in O Opm  concentrations  from  the
pile. A possible explanation for this phenomena  is that the wind
erosion emission rate of particles  < 10>im is fairly constant at
windspeeds above  a threshold of about a 7 miles/hr hourly aver-
age.  Additional soil  losses  associated with higher windspeeds are
particles >10jim not detectable by the RAM-1.
   In  tests  of the pile  with a chemical dust suppressant  applied
within three days of testing, control effectiveness in the range of
50%  was measured. Adding the windscreen to a pile previously
treated with the dust suppressant showed no incremental increase
in control effectiveness.
   In tests with different pile to windscreen distances, it was found
that a distance of two screen heights was superior to a distance of
five to eight screen heights. In fact, with the same pile height as
screen height, concentrations at 50-80 ft from  the pile were higher
at a 6.6 ft height with the screen than without the screen. This is
caused by wind shear from the screen.
Control of Dust Emissions from Active Site Cleanup
  The  operation  selected  for testing  consisted  of  a  front-end
loader (PEL)  and  dump  truck combination.  The PEL  scraped
material from the  surface, turned, traveled to the  dump truck,
and dumped its  load into  the truck. This activity simulated the
most  common method of  loading contaminated soil into trucks
for off-site disposal.
  Rather than adapting a  test  array to  the  requirements of an
existing cleanup site operation, it was  decided  to  use  captive
equipment  at a non-hazardous site. The equipment activity could
be directed by the  Held team to fit the needs  of  the testing with-
out interfering with a production schedule and without other dust
sources interfering with the testing. The use of a  noncontami-
nated site for  testing required the  assumption that the  toxic ma-
terial  would be uniformly dispersed in the soil, and that toxic soil
particles and non-toxic soil particles behave the same in the air.
  The exposure profiling  method  was used to sample the  dust
emissions downwind of the operation.  This  method employs a
tower  with  multiple  profiling  heads to  perform simultaneous
multipoint isokinetic sampling over the plume cross-section.  This
technique is applicable to point and line sources where a ground-
based profiling tower can  be located across the plume  cross-sec-
tion and where the distance from the source to the samplers can
remain  fixed. The primary sampling instruments were profiling
heads utilizing the  stacked  filter concept.  The sampling head in-
ternally fractionates  the dust  sample by particle size. Multiple
heads were mounted on each tower to sample at  several points in
the plume  directly downwind of the operation.  Three profiling
towers were used.  One tower was  located upwind of the opera-
tion to measure the background concentration not attributable to
source operation. A similar sampling tower  was located down-
wind of the PEL scraping  and traveling path  to test the scraping
emissions. The other tower was equipped with a  horizontal cross
arm and was located downwind of the dump truck to  test the
dump cycle emissions.
  Four  control measures were tested. Only two are reported in
this paper due to data availability.  On each sampling day,  both
an uncontrolled and  controlled test were  made. The control effi-
                                                        ciency is merely the difference in emission rate per unit activity
                                                        between the uncontrolled and controlled tests.
                                                          Control  Measure  I  consisted of spraying the active working
                                                        area of the PEL and dump truck with water. In the few instances
                                                        where dust control measures are currently in use at hazardous waste
                                                        sites, this is the control being used. Using a portable 200 gal tank,
                                                        pump, generator, hose and nozzle, the working area was sprayed
                                                        prior to the controlled test and again during the test as the field
                                                        team noted drying of the surface and visible emissions. Watering
                                                        amounts averaged 0.9 gal/yda. More water was used on the travel
                                                        paths of the PEL and dump truck than on the scraping area. These
                                                        active travel areas dried much faster than the scraping area. It was
                                                        noted throughout the testing that only the top 1  to 2 in. of the sur-
                                                        face were dry. Below this dry crust, the soil was very damp.
                                                          Por Control Measure 2, application procedures were identical to
                                                        those used in plain water application. However, a surfactant, John-
                                                        son-Marsh Compound MR,  was added to the water to form a
                                                        1:1000 dilution of surfactant to water. Somewhat less watering was
                                                        needed for these controlled tests. The surface remained damp
                                                        longer, and fewer subsequent applications were  required. Applica-
                                                        tion of the water/surfactant mixture averaged about 0.15 gal/yd'.
                                                          Results  are shown in Table  3. Emission rates for PEL travel
                                                        and scraping are in units of Ib/vehicle mile traveled. Emission rates
                                                        for the PEL dump are in units of g/yd'. Emissions rates can be used
                                                        to  approximate off-site exposure with assumptions about panicle
                                                        deposition. Water controlled PEL travel/scraping emissions  by
                                                        42% and 64%,  respectively, for <30-um and f 2.5-um (TSP and
                                                        PP)  size particles. Surprisingly, the emissions  from  the material
                                                        dump were reduced 63 and 70% for TSP and FP micrometer size
                                                        particles. Adding surfactant to  the water increased control effic-
                                                        iencies slightly,  while allowing the quantity of water used to be re-
                                                        duced. TSP control efficiency  for the PEL travel/scraping  in-
                                                        creased from 42% to  69%  with the addition  of the surfactant.
                                                        Other control values were similar.
                                                                                   Table 3
                                                         Control Efficiency of Area Spraying for Controlling Emissions from Soil
                                                                          Loiding by FEL Into • Track
Control and
Index
Emission Rates*
No Contra)
< 2.5 um «_ 30 urn
Control
i 2.5 um
«. 30 urn
Control
Efficiency.
i 2.S IA
I
<30^
                                                                                Front-end Loader Travel/Scraping
water (5 tests)
Ka«. Value
Mm, Value
Mean Value
Surfactant (4
tests)
Mai. Value
Min. Value
Mean Value

0.96
0.09
0.48


0.79
0.20
0.48

LSI
0.41
1.11


2.69
1.16
2.01

0.28
0.03
0.1S


0.29
0.01
0.1?

0.86
0.30
0.62


0.66
0.4S
O.S9

90
30
64


95
45
66

S6
27
42


75
61
69
                                                                                 Front-end Loader Material Dunp
Water (5 tests)
Ma«. Value
Min. Value
Mean Value
Surfactant (4
tests)
Man. Value
Min. Value
Mean Value

1.18
O.OS
0.74


0.8S
0.17
0.59

3.24
0.26
2.18


8.98
4.36
6.88

0.33
0.03
0.10


0.43
0.07
0.22

1.42
0.15
0.6S


2.93
0.94
LSI

95
40
70


82
46
62

88
31
63


88
56
77
                                                       ' Emission rates for FEL travel/scraping are in units of Ibs/vehicle nile
                                                         traveled.  For the FEL material dump, emissions are in units of grams/yd*.
268
SITE SAFETY

-------
                     SAFETY  PLANS  FOR UNCONTROLLED
                                HAZARDOUS WASTE SITES

                                               J. ROBERT STEELE
                                       Chemical Waste Management, Inc.
                                                Oak Brook, Illinois
INTRODUCTION
  The  development  and implementation  of effective,  workable
safety plans for investigation and remediation projects at uncon-
trolled  hazardous waste sites can be challenging tasks for even the
most experienced health and safety professional. These operations
can present an alarming array of potential chemical and physical
hazards that must be recognized and controlled for worker protec-
tion. Ranging  from  the multiple inherent hazards of the chem-
icals to physical hazards from  earth moving, excavation, trench-
ing, demolition and heat stress, the potential risks to workers de-
mand a comprehensive detailed site safety plan. Key elements con-
sidered essential for meeting this challenge are presented as a basic
framework for developing such a safety plan.
Safety Plan Components
  Each site safety plan should include, as a minimum, the follow-
ing categories of information presented in sufficient detail to ade-
quately address specific site hazards:
•Introduction, purpose and  scope
•Key personnel and assignment of safety responsibility
•Job hazard analyses
•Air Monitoring requirements
•Medical surveillance program
•Employee training and information
•General safe work practices
•Personal protective equipment
•Work zone delineation and decontamination procedures
•Emergency response plans
•Site security measures
•Recordkeeping requirements
  The contents of each plan section will vary depending upon the
scope of the site operations, and individual  safety plans may re-
quire additional sections. For example, maintaining an up-to-date
version of a plan once it has been approved and implemented is
extremely important. Some planners may wish to include a section
that specifically describes the mechanism  for amending the pre-
liminary safety plan. The author's intent is to provide a general out-
line, to list important considerations for each section and to pro-
vide sufficient information  for  development  of a workable, com-
prehensive safety plan.
biological hazards likely to be encountered; the overall intent of
the plan; and the scope of authority with regard to contractors,
subcontractors, site visitors and regulatory agency personnel.
  Written safety procedures for site operations must be based on
the best available information. For preparing the preliminary site
safety plan, historical information describing past cleanup opera-
tions, previous waste streams, off-site releases, news media cover-
age, recorded health or safety violations, etc., can be very useful
tools for the industrial hygienist or safety professional. Useful
sources of historical information might include:
•Site records such as waste receipts, storage inventories, manifests
 or shipping papers
•Waste generator records from firms that contributed waste to the
 site
•Regulatory agency records such as those from local or regional air
 and water pollution control boards,  OSHA, fire departments,
 health departments and federal or state investigative teams
•Previous site workers, management personnel and residents near
 the site who may have personal knowledge of site operations'
  When actual operations begin, the preliminary safety plan must
be amended, enhanced or rewritten to provide operational safety
procedures. An initial site survey should be conducted prior to this
stage to determine the actual extent of chemical, biological, radio-
logical and physical hazards likely to be encountered by site work-
ers. The site survey should include both a walk-through survey of
the work areas and initial monitoring and sampling of site mater-
ials. Information from the site survey should be readily available
to site workers and may even be included as an integral part of the
safety plan. More commonly,  however, this information is pro
vided in crew briefings, agency meetings and site log books.
KEY PERSONNEL AND ASSIGNMENT OF SAFETY
RESPONSIBILITY

  The plan should include a complete listing of key supervisory,
management and safety personnel and a description  of their re-
sponsibilities for implementation of the plan. Clear lines of author-
ity must be established for enforcing compliance with the safety
procedures. Responsibility must be assigned for  determining the
existence  of unsafe conditions,  and proper authority must  be
granted to stop unsafe operations.
INTRODUCTION, PURPOSE AND SCOPE

  The introductory paragraphs of the safety plan should describe
the nature of the work planned; potential chemical, physical and
JOB HAZARD ANALYSIS
  The job hazard analysis is used to identify those job categoric1
with the greatest hazard potential and to aid in specifying the per
                                                                                                 SITE SAFETY
                                                       269

-------
sonal protective  equipment  selection  procedure.  The analyses
should describe each job function, provide a qualitative estimate of
the employee exposure  potential  for each job function and  de-
scribe the expected production rates and worker efficiency for each
operation.2
  The analyses should also include a complete listing of expected
chemical hazards with permissible exposure limits or recommended
threshold limit values. Recommendations for specific personal pro-
tective equipment  and protective measures should be included in
this section of the plan.
  In addition to  the  chemical  hazards, consideration  should  be
given to physical  hazards such as being struck by,  struck against
or being caught in, on or between site materials, structures or ma-
chinery. The analysis  should include  such items as: slipping, trip-
ping and falling hazards, work at elevated locations,  heat stress, ex-
cessive noise, confined space work and electrical hazards, to name
a few.

AIR MONITORING REQUIREMENTS

  Specific  air monitoring procedures, including collection meth-
ods, sampling strategies, recordkeeping and action levels for imple-
menting contingency plans, both on and off the site, should be de-
veloped for each site safety plan. The detail required  for this sec-
tion of the plan is dependent in most cases on the nature of the  site
contaminants, wind and weather conditions, the proximity of the
site in relation to homes and businesses  and the specific regulatory
agency requirements. There are, however, basic components of an
air monitoring program that deserve consideration in any plan.
Buecker1 lists the following objectives for site air monitoring:
•Monitor for excursions above on-site and perimeter action levels
 for the purpose of mitigating emissions or initiating evacuation
•Substantiate the selection and use of appropriate levels of respir-
 atory protection and protective clothing
•Provide a continuous historical record of personal exposures and
 site emissions, including baseline emissions
•Establish a sample characterization mechanism to be  used for the
 screening of samples or for a contingency such as  a chemical ex-
 posure incident.

MEDICAL SURVEILLANCE

  Initial medical  surveillance for hazardous waste site workers
should include:
•Thorough review of the employee's medical, personal, family and
 occupational histories
•Thorough physical examination and clinical evaluation
•Laboratory evaluation for selected biological samples  such  as
 blood chemistry,  heavy metals screen, cholinesterase levels, etc.,
 at frequencies determined by the physician*
  The safety plan should provide sufficient  information to  site
management and medical personnel to make valid decisions about
medical acceptability of assigned workers. In most cases, the med-
ical surveillance program should be administered by a qualified  site
industrial hygienist or safety professional.
  Ongoing medical surveillance and biological sampling criteria
should be provided by an occupational or industrial physician.  In-
formation on specific medical examination requirements should be
listed in the safety plan.

EMPLOYEE TRAINING AND INFORMATION

  Before an employee can be expected to work safely, he or  she
must be properly trained. The safety plan should at least describe
the minimum acceptable  level  of safety  training required  for
assignment to a  hazardous waste site.  The purpose of the basic
safety training is, of course, to make employees aware of the safe-
ty and  health hazards they will encounter, the procedures and
equipment required to  protect  themselves and their  role during
emergency conditions. The extent of required training will depend
greatly  on  the specific job  assignment  and information gathered
from the job hazard analyses.
                                                          Certain hazardous operations will require additional  training
                                                        beyond  the traditional employee orientation, hazards awareness
                                                        and respirator training. Some specific examples of operations that
                                                        require additional employee training include: confined space en-
                                                        try, explosives handling, welding and cutting operations, emer-
                                                        gency first aid and CPR and fire fighting.
                                                          Employee information sources should also be addressed in the
                                                        site safety plan. This might include a description of the location of
                                                        material safety data sheets, air sampling results, appropriate health
                                                        and safety posters, technical reference texts and names of person-
                                                        nel with specific site safety information.

                                                        GENERAL SAFE WORK PRACTICES

                                                          This section should describe  the general measures workers must
                                                        take  to prevent exposure to hazards. Examples of these measures
                                                        would include:
                                                        •Exclusion of food, beverages and tobacco products from the con-
                                                         taminated work areas
                                                        •Personal hygiene requirements include the need for showering at
                                                         the end of the work shift, washing the hands and  face before eat-
                                                         ing, drinking or smoking
                                                        •Labeling requirements  for containers of debris,  waste, cleaning
                                                         materials, etc.
                                                        •Segregation procedures for reactive and incompatible waste ma-
                                                         terials
                                                        •Precautions which may require additional,  more specific safety
                                                         procedures for operations such as excavation and trenching,  con-
                                                         fined space work, hot work, explosives handling, etc.
                                                          One particular safety concern that should be addressed as part of
                                                        the general safe work practices is the need for  heat stress control
                                                        measures. Hazardous waste site  cleanup operations are often
                                                        scheduled during the dry summer months to avoid rainwater and
                                                        groundwater accumulation problems. Workers, equipped with im-
                                                        permeable vapor barrier clothing and respirators, are at greater risk
                                                        from heat stress and the associated heat illnesses.'
                                                          Several different heat stress indices have been developed through
                                                        the years as researchers  sought to find an  index which is physio-
                                                        logically valid over a wide range of hot environments. More than a
                                                        dozen heat stress indices are described in the research literature.*
                                                        The most commonly referenced index, the Wet Bulb Globe Temp-
                                                        erature (WBGT), has been widely accepted and is often  used to
                                                        develop  work/rest schedules for hazardous waste site operations
                                                        to control heat stress.
                                                          The safety plan might also  include some  of the USEPA  heat
                                                        stress monitoring  criteria' such as:  measurement of before  and
                                                        after work body weight  to ensure adequate body fluid replace-
                                                        ment; monitoring  of the heart rate during rest periods to ensure
                                                        proper recovery from heat stress; and monitoring of oral tempera-
                                                        tures (as an indicator of core temperature), also to ensure proper
                                                        recovery during rest periods.
                                                          Consideration should also be given to the recommendations for
                                                        adequate heat acclimatization described in the NIOSH criteria doc-
                                                        ument for heat stress control.'

                                                        PERSONAL PROTECTIVE EQUIPMENT

                                                          This section of the safety plan must describe specific respiratory
                                                        protective devices and  protective apparel  required for each job
                                                        classification and/or specific operation on-site.  The established
                                                        USEPA  protection  levels  for personal  protective  equipment,'
                                                        coupled  with an ongoing assessment of both respiratory and  skin
                                                        hazards,  can serve as  a basis for developing site  specific protec-
                                                        tion levels for the safety plan.
                                                          The USEPA protection levels describe personal protective equip-
                                                        ment requirements in four categories according to the degree of
                                                        protection afforded. These are:
                                                        •Level  A should be worn when the highest level of respiratory,
                                                         skin and eye protection is needed. Equipment would include an
                                                         SCBA, a fully encapsulating  environmental suit and appropriate
                                                         head, hand and foot protection.
270
SITE SAFETY

-------
•Level B should be worn when the highest level of respiratory pro-
 tection is needed, but a lesser level of skin protection can be safe-
 ly used. Equipment would include an SCBA, appropriate chemical-
 resistant clothing (selection based on types of exposures, chemical
 resistance,  suit construction, permeation  protection, etc.) and
 appropriate head, hand and foot protection.
•Level C should be used where minimal respiratory and skin haz-
 ards are  present and should  be selected only when the specific
 types of respiratory hazards  are known,  the concentrations have
 been measured and the criteria for selecting air purifying respira-
 tors are  met. Equipment would include a full  face chemical
 cartridge or cannister type air purifying respirator, appropriate
 chemical resistant clothing and appropriate head, hand and foot
 protection.
•Level D is the basic work uniform and should be worn only when
 no respiratory or skin hazards are encountered.
   The OSHA respiratory protection standard (29 CFR 1910.134)'
requires that employee exposures  to harmful airborne chemical
contaminants  shall be controlled, whenever feasible, by accepted
engineering  and administrative control measures rather than rely
solely on the use of respirators for routine operations. Examples
of accepted  control measures would include the use of general and
local ventilation, enclosure or isolation of the chemical process or
operation and substitution of less toxic materials.
   Hazardous waste site operations, however, are not amenable to
the use of traditional control measures.  They frequently require
materials  handling activities with high potential  for chemical ex-
posure and necessitate the use of elaborate  levels of protective
equipment.  The selection, use and maintenance of personal pro-
tective equipment must comply in all respects to the requirements
of the  OSHA standards. Careful consideration  should also  be
given to recommendations provided  by the recognized consensus
standards for  respiratory protection such as those from the Amer-
ican  Industrial  Hygiene Association,10 the  American National
Standards Institute" and the National Institute for Occupational
Safety and Health.'2
   Specific criteria for selecting protective clothing have not yet
been developed. The American Conference of Governmental In-
dustrial Hygienists (ACGIH)  has published, however, a very use-
ful field guide and technical reference manual, Guidelines for the
Selection  of Chemical Protective Clothing,** which should aid in
the development of protection levels for the safety plan. Addi-
tional information is available from technical publications such as
the American Industrial Hygiene Association  Journal,  NIOSH
publications and the Journal  of the American Society of Safety
Engineers.
 WORK ZONE DELINEATION AND
 DECONTAMINATION PROCEDURES
   Delineation of site work zones, based on the types, locations and
 exposure potential of chemical substances, is an effective means of
 preventing or reducing the possibility of exposure and transloca-
 tion of substances. Heavily contaminated areas are delineated with
 physical barriers, control points are established and decontamina-
 tion facilities and procedures are established to prevent spread of
 contamination. The  USEPA recommends the use of three contig-
 uous work zones:
 •The Exclusion  Zone,  the area of highest contamination or with
  the greatest potential for exposure, separated from the next zone
  by the "Hotline" or step-off point (appropriate protective equip-
  ment must be donned before crossing the hotline)
 •The Contamination Reduction Zone, which provides a transition
  area between the contaminated and clean areas  and contains the
  necessary decontamination equipment, washing  areas and decon-
  tamination solutions
 •The Clean or Support Zone, the outermost area which is uncon-
  taminated  and  contains the command center  and support equip-
  ment
  The USEPA control system is based on a "worst case" situation.
Less stringent control measures can be used if timely, accurate in-
formation from air monitoring, safety inspections  and substance
and soils sampling is available.
  Decontamination procedures must be developed and included in
the safety plan. Employees must be properly trained,  and decon-
tamination solutions must be properly handled and disposed. The
extent of decontamination required will depend  on a number  of
factors, the most important being the  types of contaminants in-
volved.  Other factors may include: the amount of  contamination
on protective clothing that can be visually detected; the types  of
personal protective equipment utilized;  the specific  work function
(and hazard potential) of each individual; and the reason for leav-
ing the controlled area, either for routine or emergency exit.


EMERGENCY RESPONSE PLAN
  Emergency response (or contingency) plans can  take many forms
from simple telephone rosters, action guides and checklists to com-
prehensive, detailed procedures. Contingency plans for hazardous
waste site operations, particularly in the  remedial action phase,
should be as comprehensive and detailed as possible. Considera-
tion must be given to existing community contingency plans which
will interface with site activities. Local agencies for spill response,
ambulance, fire, police and regulatory control must  review and ap-
prove the plan. Interfacing, coordinating and using existing con-
tingency plans  should be major goals  of the site  emergency re-
sponse plan.
  Plans for emergency response should address, as a minimum,
the steps to be taken for:
•Hazardous chemical reactions, fires and explosions
•Site evacuation and emergency assembly areas off-site
•Community notification, evacuation and emergency medical
 treatment
•First aid and CPR and locations of emergency and fire fighting
 equipment
•Emergency response assistance, including ambulance,  fire, hospi-
 tal, police and poison control centers'4


SITE SECURITY

  Uncontrolled hazardous waste  sites  are typically newsworthy,
and  cleanup operations can easily become a  local attraction for
area residents. Control of access to unauthorized visitors is essen-
tial.  The security section of the safety plan should include: the spe-
cific measures for identifying authorized personnel, such as the use
of photo identification badges; a requirement for  maintaining logs
of visitor names and vehicle license numbers;  the names and spe-
cific duties of security personnel;  and the specific procedures for
controlling site access (road closures, check points, barriers, fences,
etc.). Any measures taken should be approved  and  coordinated
with local emergency service agencies such as fire, police and ambu-
lance personnel who may respond to site emergencies.

RECORDKEEPING REQUIREMENTS

  The safety plan must be well documented throughout site opera-
tions. In all cases, the records maintained must conform to estab-
lished agency policies and procedures and applicable OSHA regu-
lations. The safety plan should  describe specific documentation
requirements. Typical site documents might include:
•Employee training records including proof of respirator fit tests
•Safety equipment inspection and maintenance records
•Health and safety meeting reports
•OSHA logs of injuries and illnesses
•Accident investigation reports
•Personnel exposure monitoring records
•Emergency incident reports
•Contingency plan meeting reports
•Applicable safety regulations and guidelines
                                                                                                      SITE SAFETY
                                                         271

-------
CONCLUSIONS
  Uncontrolled hazardous waste sites pose a serious risk to workers
involved in investigation and remediation operations. Multiple haz-
ards from chemicals, structures, machinery and cleaning operations
must be controlled for worker protection.  Written, understand-
able safety procedures,  based on the best available site informa-
tion, must be provided to all personnel.
  The information provided in this paper is intended to serve as a
basic outline for the development of effective safety plans.  In any
situation, the extent and application of the plan sections presented
should reflect the special considerations of each specific site.
REFERENCES

 1. PEDCo Environmental, Inc., Hazardous Waste Industry Self-Evalua-
    tion Instrument: A Health and Safety Checklist.
 2. Summers, R.L., "Analyzing Jobs," Professional Safety, 27, Dec.,
    1982, 20-23.
 3. Buecker, D.A. and Woods, D.B.,  "Air Monitoring  Strategies for
    Hazardous Waste Sites," American  Industrial Hygiene Association
    Conference, May, 1984.
 4. U.S. Department of Health and Human Services, A  Guide to the
    Work-relatednessof Disease, NIOSH number 79-116, Jan. 1979, 5-7.
 5.  Mihal. C.P., "Effect of Heat Stress on Physiological Factors for In-
    dustrial Workers Performing Routine Work and Wearing Imperme-
    able Vapor-Barrier Clothing", American  Industrial Hygiene Asso-
    ciation Journal, 42, 97-103.
 6.  National Institute for Occupational Safety and Health, Relationship
    Between Several Prominent Heat Stress Indices, DHEW number 77-
    109, Oct., 1976.
 7.  USEPA, Interim Standard Operating Safety Procedures, Sept., 1982.
 8.  NIOSH, A  Recommended Standard for Occupational Exposure to
    Hot Environments, HSM number 72-10269.
 9.  OSHA, Code of Federal Regulations, Title 29, Section 1910.134.
10.  American Industrial Hygiene Association, Respiratory Protection, A
    Manual and Guideline, AIHA, Akron, Ohio, 1980.
11.  American National Standards Institute, Inc., Practices for Respira-
    tory Protection. ANSI Z-88.2-1980, New York, May-Feb.. 1983, p.
    3-5.
12.  NIOSH, A Guide to Industrial Respiratory Protection, DHEW Num-
    ber 76-189, 1980.
13.  American Conference of Governmental Industrial Hygienists, Guide-
    lines for the  Selection of Chemical Protective Clothing, Cincinnati,
    OH, 1983.
14.  Federal  Emergency Management Agency, Planning Guide and Check-
    list for Hazardous Materials  Contingency Plans, FEMA-IO, July,
    1981.
272       SITE SAFETY

-------
         HEAT STRESS  MONITORING AT UNCONTROLLED
                                HAZARDOUS  WASTE SITES

                                         RODNEY D. TURPIN, ERT

                                       WILLIAM KEFFER, Region VII
                                            CHRIS VIAS, Region IX
                                    MARY HELEN WORDEN, Region VI
                                    U.S. Environmental Protection Agency
                                                   JOHN KING
                                                  Tetratech, Inc.
                                            San Francisco, California
INTRODUCTION
  It was not the intent of the authors of this paper to discuss
USEPA policy or policy development. Rather, the authors wish to
share their experiences in trying to deal with hot workplace en-
vironments at hazardous waste sites and monitoring employees for
heat stress. While the USEPA, Office of Emergency and Remedial
Response, Hazardous Response Support Division, Interim Stan-
dard Operating Safety Guides is not USEPA policy, it is used by
many USEPA personnel as a basic reference in addressing such site
specific problems.
1982 INTERIM STANDARD OPERATING
SAFETY GUIDES
  The Interim Standard Operating Safety Guides, published in
Sept. 1982, while not USEPA policy, suggests that there be four
primary classifications for protective clothing and respirator pro-
tection:
•Level A—Encapsulating suit with SCBA
•Level B—Hooded Chemical splash suit with SCBA
•Level C—Hooded skin protection  with air purifier canister
 respirator
•Level D—No respirator protection and minimum skin protectors
  Since the Guides discourage the use of Level D, site activities are
conducted in Level A, B or C, thus  presenting a potential for heat
stress during warmer periods. In addition, Regions II, IV, VI and
IX find themselves with the additional problem of working in areas
where temperatures and humidity are higher than generally found
within other Regions in the continental United States. .
  The Guides suggest a heat stress monitoring program be  im-
plemented when employees are wearing impervious clothing and
ambient  temperatures are 70 °F or above.  The  frequency  of
monitoring should increase as temperatures increase and employees
should be monitored after every work period once temperatures ex-
ceed 85 °F.' The Guides suggest the following monitoring program:
•Heart rate (HR) should be measured by the radial pulse for 30 sec
 as early as possible in the resting period. The HR at the beginning
 of the rest period should not exceed 110 beats/min. If the HR is
 higher, the next work period should be shortened by 10 min (or
 33%), while the length of the  rest period stays the same. If the
 pulse rate is ! 00 beats/min at the beginning of the next rest period,
 the following work cycle should be shortened by another 33%.
•Body temperature should be  measured orally with a clinical
 thermometer as early as possible in the resting period. Oral tem-
 perature (OT) at the beginning of  the rest period should not ex-
 ceed 99 °F. If it does, the  next work period should be shortened
 by 10 min (or 33%), while the length of the rest period stays the
 same. And, if the OT exceeds 99.7 °F at the beginning of the
 next period, the following work cycle should be further shortened
 by 33%. OT should be measured again at the end of the rest per-
 iod to make sure that it has dropped below 99 °F.
•Body water loss (BWL) due to sweating should be measured by
 weighing the worker in the morning  and in the evening. The
 clothing worn should  be  similar at both weighings; preferably,
 the worker should be nude. The scale should be accurate to plus
 or minus 0.25 Ib. BWL should not exceed 1.5% of the total body
 weight. If it does, the worker should be instructed to increase his
 daily intake of fluids by the  weight lost. Ideally, body  fluids
 should be maintained  at a constant level  during the work day.
 This  requires replacement of salt lost in sweat as well.
•Good hygienic standards must be maintained by frequent change
 of clothing and daily showering. Clothing should be permitted to
 dry during  rest periods. Persons who notice  skin problems
 should immediately consult medical personnel.
  While these guidelines are extremely helpful,  it is not always
possible to implement them at every hazardous waste site. For ex-
ample, the  Environmental Response  Team  (ERT), as  well as
various USEPA Regional personnel, have found it extremely dif-
ficult to obtain body weights on-site. ERT has found it helpful to
substitute taking the employee's blood pressure for body weight.
While this is not a direct measurement for heat stress monitoring, it
does  alert  on-site  safety and health  professionals of  those
employees who should be watched closely because of high  blood
pressure as well as  indicate an individual's overall stress as the
response develops.
CASE HISTORIES
Dioxin Site, New Jersey

  Site activities occurred during a typical hot, humid New Jersey
summer day. What makes operating procedure here noteworthy is
some of the heat stress equipment utilized during this activity. The
site was located within an industrial park in central New Jersey. It
consisted of several large concrete slabs which were once the floors
for warehouses,  etc., broken asphalt parking lots, storm drains,
sand/clay/gravel spots as well as some vegetation areas.
  All site personnel were wearing  Level C (Hooded Tyvek/Saran
disposable suits, full face air purifying respirator with industrial
size pesticide canisters, gloves, etc.). The on-site tasks consisted of
two major efforts: collecting and blending soil samples.
                                                                                              SITE SAFETY
                                                    273

-------
  The blending operation required the use of 110 A/C power and
was located near the decontamination area. Since shade trees were
absent,  inexpensive beach umbrellas were  purchased and placed
within the area to extend the work periods from the scheduled 20
min. Blender personnel were in the same Level C protection as the
samplers: however, they did not have communication radios.
  The  soil samplers  were in  Level  C protective equipment  and
worked for 20 min followed by a 20 min rest period. The sample
grid had been laid out in advance to reduce the amount of time on-
site. The decontamination line was modified slightly from those set
forth in the Interim Standard Operating Safely Guides by adding
an annex. The annex, designated as a rest area,  was located  just
past the gross decon area in the decon line. This area was shady and
stools were provided.
  In attempting to implement heat  stress monitoring as described
in the Guides, the following were some of the major problems en-
countered:
•The accuracy of the scale was inconsistent.
•Blood pressures were  hard  to hear and a  limited number of
 people were qualified to take them.
•The decon line was slowed due to the timely process of taking
 temperatures.
•It was difficult for the site  Health and Safety Officer to take
 pulses at the  appropriate time.
  The following is a summary of how the problems were addressed
and resolved:
•Since the duration of the on-site sampling project was less than
 6 days, the body weight requirement was deleted from the on-site
 monitoring. Blood pressure measurements were taken at each rest
 period instead of the body weight.
•Initially, the blood pressure was  obtained with the  traditional
 stethescope  and sphygmomanometer. However, this had  two
 major drawbacks:
   1. The Safety and Health  personnel had a difficult time hear-
      ing blood pressure because of background noises, i.e., diesel
      engines and employee conversations.
   2. In addition, it took approximately 3 to 8  min to  monitor
      each person coming into the rest area. This type of monitor-
      ing became a "bottle neck"  in personnel traffic flow.  The
      problems were eliminated by incorporating the use of three
      digital electronic blood  pressure monitors.
•Oral temperatures were originally  collected  using  a  traditional
 oral thermometer and this slowed personnel  movement  through
 the decon line. This problem  was eliminated by the use of several
 digital thermometers for the sampling crew to use during their
 rest period.
•Pulse  rates were taken at the wrist and/or throat. This was a time
 consuming and difficult task to conduct on personnel coming
 off  the site. This problem  was solved since the  digital electronic
 blood pressure monitors also took  pulse rates.
 Pacific Islands

   The project began in January 1983 when the Region IX, USEPA
 Emergency Response Section and the Technical Assistance Team
 (TAT) were asked to perform preliminary assessments on sites in
 Micronesia. The sites were located in Guam, an unincorporated ter-
 ritory of the United States,  and  several islands in the Trust Ter-
 ritories of the Pacific  Islands (TTPI).  The TTPI consists of  the
 emerging political entities of the  Commonwealth of  the Northern
 Marianas  Islands  (CNMI),  the  Federated  States of  Micronesia
 (Kosrae,  Ponape, Truk and Yap), the  Republic of  the Marshall
 Islands and the Republic of Palau. Guam, the center of business ac-
 tivity for the region, is located approximately 2700  km south  of
 Tokyo and 500 km west of Honolulu.
   Approximately 50% of the sites contained improperly stored and
 leaking transformers. Previous site surveys had identified many of
 the transformers  as being PCB-contaminated. The  remainder  of
 the sites primarily contained pesticide and miscellaneous hospital
 laboratory wastes.
                                                         One of the prime concerns during the planning stages was the ef-
                                                       fect of the extreme temperatures and  humidity. The Islands are
                                                       located only 7 to 13 degrees above the Equator and temperatures
                                                       range  between  85° and 100°F even  during the  night. Humidity
                                                       averaged from 80 to 100%. Since the team would be wearing pro-
                                                       tective gear, the heat stress potential was substantial.
                                                         In addition to heat-related illnesses, heat  stress can increase the
                                                       probability of accidents and result in a loss of efficiency. This was
                                                       an important factor to consider since each team operated with only
                                                       six  people. Although  an adequate amount  of time  was initially
                                                       allocated for each  island cleanup, operational delays coupled with
                                                       the discovery of additional sites,  required  the team to work 18
                                                       hours a day.
                                                         The team completed 35 cleanups in 12 weeks. The majority of the
                                                       work, which required protective gear, was carried out after sunset.
                                                       This was the most important factor  in reducing the heat stress
                                                       potential. In addition,  five other factors warranted attention.

                                                       Choice of Monitoring Guidelines
                                                         The existing USEPA and TAT Region IX heat stress policy calls
                                                       for monitoring  oral temperature, blood pressure and  pulse rate in
                                                       addition to weather conditions using an electronic thermometer
                                                       and sphygmomanometer,  an  outdoor thermometer  and a wind
                                                       speed indicator. This policy was adopted in the Pacific cleanup.
                                                         Once in the  Islands, a  USCG  Emergency Medical Technician
                                                       monitored each person for conditions which could  increase the
                                                       probability of heat stress;  however, it  was occasionally necessary
                                                       for workers to monitor themselves. Workers were more concerned
                                                       with cooling down (drinking liquids, stripping protective clothing)
                                                       than heat stress monitoring and this decreased the ability of the
                                                       program to identify heat stressed workers.
                                                       Availability of Emergency Medical Care
                                                         Only basic emergency medical care was available on each Island
                                                       and the team faced communication difficulties as each Island had
                                                       its own native language. For these reasons, it was necessary to make
                                                       provisions for  the airlifting of injured  personnel to  Guam or
                                                       Honolulu. Unfortunately,  air service  to the  Islands is limited,
                                                       averaging only two to three flights per week. In addition, suitable
                                                       charter aircrafts were unavailable on the outer Islands. Therefore,
                                                       the team  made  arrangements  with the Department  of Defense,
                                                       which operates Civilian Assistance Teams (CAT Teams) on most of
                                                       the Islands, to provide medical air evacuations if they were needed.
                                                       Availability of Potable Water
                                                         It was expected that potable water would be unavailable on most
                                                       Islands since Micronesia was suffering a severe drought. The teams
                                                       also anticipated thai the water available would be contaminated
                                                       with bacteria. This was compensated for by equipping both teams
                                                       with halozone, a chlorine based disinfectant. Because this chemical
                                                       leaves an unpleasant taste in the water, the team made fruit juices
                                                       available. As it  turned out, water was available almost everywhere
                                                       but was usually contaminated.  On some Islands, the team was able
                                                       to obtain potable water from the CAT Team.
                                                       Acclimatization

                                                         Climate acclimatization was a major problem in the response ef-
                                                       fort. Because the team was used to the generally cool San Francisco
                                                       climate, it required up to a week for all of the team members to ac-
                                                       climate to the Micronesian heat and humidity. Temperatures and
                                                       humidity conditions also varied from  island to island, sometimes
                                                       significantly. This variation necessitated an additional acclimatiza-
                                                       tion period of several days.  Unfortunately, because interisland
                                                       fiights were infrequent, these periods were sometimes  unavailable.

                                                       "Cool Down"
                                                         The team's concern with "cool down" was twofold. The proba-
                                                       bility of heat stress can be decreased by removing workers to a cool
                                                       environment following  work periods.  When cleanups  were in pro-
                                                       gress, the teams frequently worked 18-hr days so there was little
                                                       lime to  recover from one day's  work  before  another began.
274
SITE SAFETY

-------
Second, during mainland responses where heat stress problems are
anticipated, firetrucks can be used to spray workers after their en-
tries. This was not possible in the Islands, but ocean water and
tropical rains adequately compensated for this. Air conditioning
was the exception throughout Micronesia.
  Generally, responses can be carried out safely in hot and humid
remote areas of the world as long as the planning is adequate. From
the USEPA's experience in the Islands and elsewhere in the region,
the authors believe that blood pressure is clearly the most variable
and least reliable indicator of heat stress. Oral temperature and
heart rate are much better indicators. Both these parameters cor-
related with high temperature and humidity as well as with com-
ments team members made about how they felt.
 Triangle Chemical Site

   This chemical company site near Orange, Texas, abandoned in
 1981, was identified  as  a significant environmental and public
 health hazard. A planned removal action was carried out in August
 1982. The ambient temperatures ranged from 86 °F to 100 °F with
 humidities between 80 and 100% during the entire operation.
   The site safety  plan included provisions  for monitoring heat
 stress of employees.  Pre-work  pulse  rates  and blood pressure
 measurements  were  recorded during morning safety  meetings.
 Readings were taken again after each on-site visit.
   Workers were instructed on the importance of water balance to
 persons  who  perform tasks in hot  environments wearing  im-
 permeable clothing while performing  strenuous activities. They
 were advised to increase their daily intake of fluids to offset body
 water loss due to sweating and to increase the amount of salt used
 on food. Suitable  replacement fluids were made available during
 working hours.
   Body  temperatures and weight loss were not measured at  this
 location. None of the USEPA employees experienced problems
 while working. Only one individual was found to be affected by an
 elevated blood pressure level.
 Cleve Reber Superfund Site
   An emergency removal action  was undertaken  at  the  Cleve
 Reber, Sorrento Site in Southern Louisiana in the latter part of July
 1983. The ambient temperatures registered 100°F daily while the
 humidity was 100%. Immediate removal activities entailed removal
 of over 500 barrels.
   The workers involved in this action were physically fit and ac-
 climated to working in hot environments. Pulse rates were recorded
 at between 70-80 at rest.  However,  some individuals had an in-
 crease in heart rate of 180-190 beats per minute within 10 minutes
 after work began.  Sensing thermometers were  worn  inside the
 workers' clothing; readings reached 125 ° while working in the sun
 and lowered to around 100 °F when in the shade.
   The initial work schedule called for 60 min of work without stop-
 ping. The  next hour consisted of 45  min work with  a 15 min rest
 break. By 1:00 p.m., the schedule was 30 min work and 30 min rest.
 Eventually, the  personnel were unable to work for more than 20
 min at a time because of the intense heat and 20-340 min were re-
 quired for recovery.
   Portable showers were constructed using large fiberglass tanks
 for water storage,  a shower head and  a  portable  electrical
 generator. Workers moved to the established clear area, removed
 their Tyvek suits, gloves and tape, showered down in their clothing,
 sat down  on a bench to rest, suited back up again with  wet
 underclothing and  returned for another  work period. This pro-
 cedure provided a rapid cooling down method.
   Overall observations noted included:
 •USEPA workers could not work as  long as contractor personnel.
  The reasons included such factors as Saran vs. plan Tyvek suits,
  completely taped out workers around the glove and boot areas
 vs. those that were not and full face respirators vs. 1/2 masks on
 contractor personnel.
•Oral temperatures  were not  always reliable.  Some  workers
 registered  103 °F  temperatures  with  no  apparent stress while
 others had readings  of 97 °F with numerous effects.
•Half-mask respirators with faceshields and goggles were much
 easier and more comfortable to work in than full-face respirators.
•PVC or Tyvek/Saran suits are much  more uncomfortable when
 working in hot environments than Tyvek suits. Plain Tyvek suits
 should not be worn unless chemical splash is not considered to be
 a problem.
Eastern Missouri

  Since November  of 1982, seven investigatory  phases of the
eastern Missouri dioxin investigations have been completed. The
project  involved the  sampling of soil  at  more than  100 sites
throughout eastern Missouri and the collection of nearly 5,000 soil
and water samples for 2,3,7,8-TCDD  analyses. The sampling pro-
cedures required the use of hand augers, picks, shovels and drill
rigs.
  The personnel protective equipment required while sampling for
dioxin includes neoprene steel-toed work boots, full-body one-
piece  impervious protective  suits,  internal  gloves  of lightweight
vinyl or latex, external viton  gloves and a full-face respirator with
combination organic vapor high-efficiency  particulate cartridge/
canisters.  Wearing this type of personnel  protective equipment
poses a  problem of possible heat  stress as  the ambient air
temperature increases.
  The work load of the personnel performing the sampling was
probably moderate to heavy, and  work at  times involved  lifting
heavy objects such as augers, as well  as strainful pick and shovel
work.  The work at  other  times  was less vigorous, involving
moderate walks or spooning soil into sample jars. When consider-
ing the type of work load involved, the permissible heat exposure
TLV's are only valid when  the worker is wearing light  summer
clothing. Since the clothing worn by the personnel was cumbersome
and nearly impermeable, the  work load was considered heavy.
  The safety plan required that a heat stress monitoring program
be performed as outdoor temperatures increased. The team  leader
calculated the WBGT at two hour intervals when the daily max-
imum  temperature  exceeded 60 °F.  When the WBGT  reached
82.2 T  or higher,  medical surveillance  of  pulse  and  body
temperature of team members were initiated.
  The ambient outdoor temperatures were recorded upon arrival to
the site and continually noted each hour until the on-site work was
completed. Individual monitoring of those employees who were
fully suited up (Tyveks, full-face respirators, viton gloves and steel-
toed  boots) also took  place throughout the day.  Personnel
monitoring included weighing in  and out each day,  along with
periodic checks of employees oral temperatures, pulse rates and
blood pressures.
  The equipment used for health monitoring included a bathroom
scale  for  weighing the employees,  a sphygmomanometer  for
measuring blood pressures and two IVAC monitors for determin-
ing employee's temperatures and pulse rates. A battery operated
WBGT heat stress monitor was used  for monitoring the  outdoor
temperatures.
  Personnel monitoring began in the morning before leaving the
Command Post. Each employee's weight was recorded along with
his temperature, pulse rate and blood pressure. Upon arrival at the
sampling site, the team set up its WBGT heat stress monitor and
recorded the dry bulb, wet bulb,  globe  and wet  bulb  globe
temperatures.  As  the   employees  began  suiting   up,   their
temperatures, pulse rates and blood pressures were measured. This
personnel monitoring was repeated each  time the employees un-
suited for a break (approximately every one to two hours). After
returning  to  the Command Post,  the  employees  were  again
weighed to determine the amount  of weight lost throughout the
day.
                                                                                                     SITE SAFETY
                                                        275

-------
   No  employee  experienced  heat  cramps  or heat exhaustion;
 however, there were a few times that workers definitely needed a
 break  in  order to unsuit and  rest. Body temperatures often rose
 above 99°F, with pulse rates of more than 100 beat/min. If an in-
 dividual appeared to be extremely hot, he was given a wet cloth to
 wipe his arms and face, Gatorade or water to drink  and was then
 seated in a shaded rest area.  All workers recovered quickly and
 were able to continue working after a 15-30 min break.

 CONCLUSIONS
   As hazardous waste site occupational health and safety profes-
 sionals know, heat stress monitoring is only one of the many dif-
ficult problems present during most operations. In addition to site
conditions, one must consider other factors  such as  overweight
workers, smokers vs. non-smokers, the food consumption of the
employees during off-duty time and the wearing of dark plastic or
rubberized items which are heated by the sun causing increased in-
cidents of burns.

REFERENCES

I.  Interim Standard Operating Safety Guides,  Revised September 1982,
   U.S.  Environmcnial  Protection Agency, Office of Emergency and
   Remedial Response,  Hazardous Response Support Division, Wash-
   ington, D.C.
276      SITE SAFETY

-------
DECISION MAKING FOR REMEDIAL ALTERNATIVES USING
 THE  PROVISIONS  OF CERCLA:  PCB RIVER CLEANUP AND
                   INDUSTRIAL SITE  CLEANUP/CLOSURE
                                        LAWRENCE M. GOLDMAN
                                             ROBERTA J. FINE
                                                  Erseco, Inc.
                                           Braintree,  Massachusetts
 INTRODUCTION

   Erseco, Inc. has utilized the evaluation criteria of the National
 Contingency Plan (NCP) of CERCLA to provide guidance and
 construction of a decision making structure during remedial action
 planning for several hazardous waste sites. By using the NCP in
 conjunction with the provisions set forth in federal and state en-
 forcement documents, Erseco was able to define the extent of
 remedy required  at these sites and to evaluate, screen and select
 remedial alternatives which  were the most cost effective and en-
 vironmentally sound.
 NATIONAL CONTINGENCY PLAN

   On December 11, 1983, the Congress enacted CERCLA, which
 establishes broad federal  authority to respond to releases or threats
 of releases of hazardous substances  by  undertaking short-term
 cleanup actions and/or long-term actions consistent with perma-
 nent remedy. Furthermore, the Congress directed  the USEPA,
 under Section 105 of CERCLA, to amend the NCP to make federal
 responses to releases or threatened releases of hazardous materials
 "reasonably predictable by both the regulated community and the
 general public".  On  March 2, 1982, the  USEPA proposed final
 rulemaking to amend the NCP which stated in the preamble that,
 based on USEPA's prior enforcement experience (pre-CERCLA),
 a flexible standard for  determining the  appropriate extent of
 remedy is the best standard  at that time,  and that a formal cost-
 benefit analysis for each remedial action alternative should be con-
 sidered. On July 16, 1982, the USEPA published final rulemaking
 amending the NCP.
   In essence, the USEPA structured the NCP provisions for deter-
 mining the appropriate extent of remedial actions by opting for a
 flexible site-by-site approach rather than  imposing rigid  national
 standards; they chose to emphasize considerations of costs; they
 also chose to ensure that all remedies under CERCLA,  whether
 publicly  or privately financed,  would be determined  through
 basically the same cost/benefit approach.
   The following case histories describe how the NCP provisions
 provided guidance  and  decision  making structure during  the
 remedial action planning for each site. The flexibility of the NCP
 criteria is evident through  its adaptation to  the particular cir-
 cumstances concerning each  case.
 #1—GROUNDWATER  CONTAMINATION
   The first example case involves the discovery of organic chemical
 contamination in the groundwater of a municipal well field adja-
 cent to a major chemical company. This discovery prompted the
 responsible chemical  company to enter into agreements with the
USEPA and the state to study the problem and begin cleanup.
These  agreements  were  in  accordance  with  the  federal
government's jurisdiction under Section 7003 of RCRA and the ap-
propriate provisions of state law. The settlement document? di-
vided the site restoration work into two major phases: (1) Site
Cleanup, and (2) Aquifer Restoration.
Site Description

  The waste disposal sites at the chemical plant consisted of three
major lagoon areas used for the disposal of process and cooling
waters and an industrial landfill used for the disposal of solid
wastes. Additionally, five "other waste sites" were used for the
disposal of waste products.
  The proximity of these sites to the municipal wells and surface
waters is shown in Figure 1. A natural groundwater divide exists
north of the manufacturing areas. To the south of the divide,
groundwater flows in a southerly direction toward the river. To the
north of the divide, the groundwater flows north toward a brook.

Contamination
  Following the discovery of contamination at the municipal wells
near the facility, an extensive groundwater monitoring network was
installed at the  site. Monitoring results indicated that several
plumes existed within the aquifer system, and the chemical com-
position of the individual plumes was characteristic of the materials
deposited at the various disposal locations (Fig. 2).
  Several halogenated hydrocarbons were consistently present, the
most common being VDC. Other halogenated compounds were
also present. Volatile aromatics such as benzene, toluene and
ethylbenzene were also found. The presence of these compounds
was consistent with known waste disposal practices of the facility.

Remedy

  According to the settlement documents agreed to by the USEPA,
the state and the chemical company, the aquifer would have to be
cleaned up and restored to a "fully usable condition."
  The term "fully usable condition" was not defined in the agree-
ments and, therefore, the required level of cleanup of the aquifer
was not specified. Similarly, the agreements did not define the ap-
propriate extent of remedy for the other waste sites. The screening
criteria used to evaluate the alternatives to  develop appropriate
aquifer restoration  and  site cleanup plans consistent with these
agreements was, therefore, complex and undefined. However, the
National Contingency Plan (NCP) provided guidance for determin-
ing the appropriate extent of remedy with respect to government as
                                                                  RISK ASSESSMENT/DECISION ANALYSIS      277

-------
                                                           Figure 1
                                              Chemical Facility Waste Disposal Sites
well as private-party financed cleanup activities. This  is clearly
stated under Section V of the March 12,  1982 Preamble to the
NCP, Enforcement Actions:
   "It is USEPA policy that the same factors used to deter-
mine the  appropriate  extent of remedy for  fund-financed
cleanup be considered  to evaluate the adequacy of or deter-
mine the level of cleanup to be sought through enforcement
efforts. Section 300.67(c) explicitly reflects this policy by pro-
viding that the criteria in section 300.67(e) through (j) will be
used  to determine the appropriate  extent of  remedy  for
private-party cleanup."
   Paragraph 0) of Section 300.67 specifically refers to the extent of
remedy as follows:
   "The appropriate extent of remedy shall be determined by
the lead agency's selection of the remedial alternative which
the agency determines  is cost effective (i.e., the lowest cost
alternative that is technically feasible and reliable) and which
effectively mitigates  the minimizes damage to and provides
adequate protection of public  health, welfare and  the  en-
vironment."
  Clearly, according to the NCP, two factors should be considered
when determining the level of aquifer restoration or individual site
cleanup required at  the facility. The  remedial alternative chosen
must  be cost-effective  and yet effectively mitigate and  minimize
damage to and provide  adequate protection  of public  health,
welfare and the environment. Using these criteria as a guideline, in
conjunction with the site-specific conditions, appropriate remedial
action plans for aquifer restoration and cleanup of the other sites at
the facility were developed to meet the requirements contained in
the agreement.
Aquifer Restoration

  Due to the existence of several plumes containing one or more
contaminants at the site, different  cleanup levels may be specified
depending upon a plume's location, flow field and contaminant
transport and behavior. From previous studies, it appeared that on-
                                                        ly a portion of contaminants in the groundwater would reach the
                                                        municipal wells. Other contaminants were flowing in the direction
                                                        of a nearby river, downstream of the municipal wells, while other
                                                        contaminants were traveling away from the municipal wells in the
                                                        direction of a brook.
                                                          Using the NCP guidelines for providing adequate protection of
                                                        public health and the  agreement  requirements  of restoring the
                                                        aquifer to a  "fully usable condition,"  the appropriate  level of
                                                        cleanup and mitigative measures for an aquifer restoration plan at
                                                        the facility would have  to comply with the following:
                                                        •Aquifer restoration intercepted by municipal wells would require
                                                         attainment of water quality  levels, after  treatment, which meet
                                                         applicable drinking water standards.
                                                        •The contaminants contained  in that part of the aquifer discharg-
                                                         ing to the nearby river could not exceed the river's capacity to
                                                         assimilate these contaminants. This assimilative capacity comes
                                                         from dilution, dispersion and evaporation.
                                                        •An Aquifer  Restoration  Program should incorporate the most
                                                         appropriate  and effective technology demonstrated for similar
                                                         cases. Such technology included the removal, treatment and sub-
                                                         sequent recharge to the  aquifer. Because no quantitative prede-
                                                         termined  levels  of cleanup had been established, both of the
                                                         previous conditions applied. In addition, the level of contaminant
                                                         removal obtainable through treatment will be determined by the
                                                         limitations of the available technology.

                                                        Other Waste Sites

                                                          The intent of remedial actions at the other sites was to control the
                                                        contamination at  the source before  it  entered the  aquifer.
                                                        Therefore, a remedial action which provided  an appropriate extent
                                                        of remedy at  these sites was one that minimized the long-term in-
                                                        filtration into the closed  site,  thereby preventing the production
                                                        and migration of Icachate. If, however, the implementation of a
                                                        remedial action caused a greater environmental or health danger
                                                        than no remedial action,  a no action alternative would  have been
                                                        appropriate.
278
RISK ASSESSMENT/DECISION ANALYSIS

-------
                                                                      North Plume
                                                                      * VDC
                                                                                      Lagoon  Plume/
                                                                                        * Ethylbepzene
                                                                                        * Benzen^
                                                                                        * Tcluepie
                                                                                       * Benzene
                                                                                       * Toluene
                                                                                       * Ethylbenzene
                                                                                       * 1,2  Dichl
    South Plume
      * VDC
               Municipal
               SSf
                                                           Figure 2
                                               Groundwater Contamination Plumes
Identification of Remedial Alternatives
Aquifer Restoration
  Several remedial technologies applicable for aquifer restoration
were identified for preliminary screening, specifically:
•Natural and physical removal of water—natural cleansing (no
 additional actions with continued pumping under present condi-
 tions) or additional recovery well pumping.
•Water treatment—activated carbon or air stripping.
•In situ Treatment—biological.
•Containment Structures—slurry walls.
Other Waste Sites
  The  specific technologies applicable for remedial action at the
other sites depended upon the nature and extent of contamination
present at the individual disposal areas. Furthermore, for the pur-
pose of remedial actions at these sites, it was decided that any off-
site remedial measures which were under consideration were more
appropriately addressed and  implemented  under the Aquifer
Restoration Plan. The intent of remedial actions at the other waste
sites was to control the contamination at  the  source  before it
entered the aquifer.  Therefore,  the  following source control
methods were identified for preliminary  screening:
•No Action
•Capping
•Encapsulation
•Containment Structures
•Incineration
•Disposal
Evaluation Criteria
  The  assessment criteria used to preliminarily screen the above
technologies were based on those specified in the  NCP (40 CFR
300.68H). This section states that "three broad criteria should be
used in the initial screening of alternatives...," cost, effects of the
alternative and acceptable engineering practices:
•Cost. For each alternative, the cost of installing or implementing
 the remedial action must be considered, including operation and
 maintenance costs.
•Effects of the Alternative. The effects of each alternative should
 be evaluated in two ways: (1) whether the alternative itself or its
 implementation has any adverse environmental effects, and  (2)
 whether the  alternative for source control remedial actions is
 likely to achieve adequate control of source material.
•Acceptable Engineering Practices. Alternatives must be feasible
 for the location and conditions of the release, applicable to the
 problem and represent a reliable means of addressing the problem.
Evaluation of Remedial Technologies:
Aquifer Restoration
  As a result of the preliminary screening process, the following
remedial actions were proposed for restoration of the aquifer:
•Physical removal
•Water treatment
•Natural aquifer cleansing
  By varying and/or combining operational modifications and dis-
charge locations of the above technologies, several remedial alter-
natives were developed. These consisted of various locations of
bedrock extraction wells in conjunction with strategically located
pumping wells and water treatment systems.
Other Waste Sites
  The preliminary screening process for those technologies  con-
sidered for the other sites eliminated containment structures and in-
cineration from further consideration. Although encapsulation and
solidification at individual sites were also eliminated, these actions
                                                                         RISK ASSESSMENT/DECISION ANALYSIS       279

-------
were evaluated further under the co-disposal option. Therefore, the
following remedial actions were proposed for the other waste sites:
•No Action
•Capping
•Disposal: on-site, off-site

Detailed Evaluation
  As part of the studies conducted at the sites,  conceptual and
mathematical computer models of the groundwater flow field and
contaminants  transport  in  the aquifer were developed. Various
pumping strategies were incorporated into the model to determine
the effects of these alternatives, which were used during the detail-
ed evaluation. The recommended remedial action  plan for aquifer
restoration is currently under government review, while the detailed
evaluation of alternatives for the other sites is ongoing.
  A detailed evaluation of the proposed technologies for the other
waste sites was required to determine which alternative would be
most cost-effective and environmentally sound.
 #2—PCB CONTAMINATION

   In the second case history to be discussed here, the criteria within
 the NCP were also effectively used during remedial action planning
 for PCB-contaminated river sediment in the northeast.
   In June, 1981, the USEPA, the state and a major industry signed
 Consent Orders which established a framework by which the com-
 pany could begin to develop a program  to conduct studies of PCB
 contamination, transport and remedial actions for the affected
 river and related bodies of water. The purpose of the Remedial Ac-
 tion Study, as stated in the orders, was  to study the relative costs,
 benefits and environmental  impacts of remedial actions or treat-
 ment of sediments so that concentrations of PCBs did not exceed
 10 and  50 ppm. In addition, the company agreed  to study the
 feasibility of the removal or treatment of sediments to PCB concen-
 trations of 1  ppm.
   Although the Consent Orders provided a broad framework of
 steps for developing a sampling and analysis program and recom-
 mended investigating specific  remedial  alternatives,  they did not
 address  the methods of response and the process for remedial ac-
 tion selection.

 Site Description

   The study area was a major river in the northeast which passes
 through several states (Fig. 3). Three  key drainage areas along the
 river basin were identified in a transport study which characterized
 three separate reaches of the river with distinctive gradients ranging
 from 0.04 to 0.2%. The significance of this change in gradient is the
 differential effect on the flow and  discharge rates for  various
 segments of the river. These changes in the flow regime translate to
 a wide  variety  of erosional, transport  and  depositional en-
 vironments for PCB-laden sediment which would, in turn, deter-
 mine variations  in PCB  concentrations and the  appropriate
 remedial action.
   One of the most significant areas of study is a pond located ap-
 proximately  18  miles from the  disposal area.  In  addition  to
 numerous shallow channels and meanders, there is a dam, a bypass
 channel  and  a   holding  pond.  The  transport  and  deposition
 mechanisms of this area  were complex because of the variety of
 flow regimes.  Deposition was compounded by several man-made
 structures which acted to  either divert sediment around the area or
 trap it.
   The stretch of river from the suspected source of contamination
 to a point  18 miles  downstream has been defined as the Remedial
 Action  Zone  (RAZ). The RAZ includes  those areas for which
 average concentrations of greater than 10 ppm occur, as prescribed
 in the Consent Orders. Each  of these  areas was, therefore, sub-
 jected to evaluation for appropriate remedial actions.
   As a result of the implementation of selected engineering  alter-
 natives,  the Remedial Impact Zone (RIZ) was defined as that sec-
                                                       tion of the river below  the  RAZ which  would be affected by
                                                       upstream remedial actions.
                                                       Extent of Contamination
                                                         The characteristic river sections of the study area previously
                                                       defined are shown  in Table  1.  For  each of these sections, the
                                                       amount of PCB, the fraction of the load and the average PCB con-
                                                       centration is given.
                                                         The average  concentration for background stations upstream
                                                       from the suspected source area was 0.15 ppm. The pond area con-
                                                       tained a variety of flow and depositional environments and had a
                                                       mean PCB concentration of 24 ppm  The range for the area was
                                                       0.09 to 100 ppm. The highest  concentrations occurred in the more
                                                       quiescent areas  of the pond along its  banks,  backwater pools, ox-
                                                       bows and behind a submerged  abutment at the head of the channel.
                                                              Background
                                                            RemsdUl
                                                             Action Zoo*
                                                                                 Figure 3
                                                                         Outline of River Basin Area
                                                         The average PCB concentration decreased to 3.1 ppm over the
                                                       next 19 mile stretch below the pond. However, this concentration
                                                       was misleading due to a series of five dams along this stretch which
                                                       trap sediment. The PCB concentrations behind two of the dams
                                                       were 0.21 ppm and 5.9 ppm; the discharge of suspended PCBs is a
                                                       factor of 10 times as great as the pond area station.
                                                         From  the data contained in  Table  1, several conclusions were
                                                       drawn. The first and most obvious was that the natural and man-
                                                       made impoundments create effective traps for sediment and PCBs.
                                                       However, discharge of PCBs was highest from impounded areas of
                                                       lower  concentration.  Therefore,  remedial  actions designed  to
                                                       reduce the PCB concentration  to specified levels in certain areas
                                                       were not necessarily the most effective in reducing the transport or
                                                       bioavailability of PCBs in the river.
280
RISK ASSESSMENT/DECISION ANALYSIS

-------
                                                           Table 1
                                                 PCB Distribution in the River
AVERAGE
PCB
PCB % OF CONCENTRATION
LOCATION SECTION LENGTH (POUNDS) LOAD (PPM)
Background
Remedial
Action
Zone
Ramadlal
Impact
Zoo*

TOTALS
1
2
3
5
6


6.33
7.8
4.4
.4
19.4
23.9

62.18
14
8,510
16,500
7,240
3,945
671

36,880
0.1
23.1
44.7
19.6
10.7
1.8

10014
.15
60
22
24
3.1
.66

—
Identification of Remedial Technologies

  The Consent Orders required the evaluation  of at least  four
remedial actions, which were:
•No Action
•Dredging and removal of PCB-laden sediment
•In situ treatment
•In situ impoundment
  Additionally, those water  diversion techniques which would
minimize and mitigate the re-suspension or transport of sediments
were  also included.  These included  sediment dispersal control
equipment  such as silt curtains  or  sheet piling and surface and
groundwater controls such as cofferdams and slurry walls.
Evaluation Criteria

  Like the previous case history, a preliminary screening of ac-
cepted remedial technologies was first required to eliminate inap-
propriate actions from further consideration.  The remaining ac-
tions were then subject to a detailed evaluation to select the most
cost-effective and environmentally sound alternative. Development
of the preliminary screening criteria was based on the three broad
criteria suggested by the NCP  and the Consent Orders—cost, ef-
fects of the alternative and acceptable engineering practices. These
criteria were used to eliminate  technologies which consisted of:
•Alternatives judged not technically feasible, including unproven
 or conceptual methods
•Alternatives not consistent with the Consent Order
•Ineffective controls in reducing exposure to  or potential health
 effects of PCBs
•Alternatives having conceptual costs  that were not cost-effective
 (i.e., costs which are an order  of magnitude higher than other
 alternatives and do  not provide a  commensurate  environmental
 benefit)
•Alternatives that required an unreasonable length  of time to im-
 plement due to regulatory requirements or  technical issues (i.e.,
 permits, pilot studies)
•Alternatives which have significant adverse environmental effects
  The selected methods were state-of-the-art technologies specific
to the  PCB  problem and  were based  on  proven application
developed for other sites with similar needs.

Evaluation  of Remedial Technologies
  The following is a summary of those techniques retained for fur-
ther evaluation after the preliminary screening:
•No  Action
•Dredging  of PCB-laden sediment using: clamshell, cutterhead
 suction, Mudcat, pneuma and namtech dredges
•Excavation of  PCB-laden  sediment using:  scraper, front end
 loader, backhoe and/or dragline crane
•Sediment dispersal control using:  silt curtains,  floating boom
 and/or sheet piling
•Surface water and groundwater control using: cofferdams, de-
 watering, sheet piling and slurry walls
•Solids dewatering  using lagoons,  drying  beds  or  container
 storage
•Chemical fixation
•In situ impoundment using: isolation (dikes, berms, bulkheads,
 sheet pilings, impermeable liners and inert  material), stabiliza-
 tion and/or channelization
•Disposal using a landfill or incineration  either on-site or off-site
•Sorbent materials
  Each one of these technologies involves  proven engineering prac-
tices which may be applicable to the river and may be effective in
reducing the PCB concentration to the desired level. However, the
following  discussion  evaluates two of the selected alternatives
which were found to  be technically feasible but required further
evaluation due to their  cost effectiveness and environmental im-
pacts.

Incineration
  Incineration of PCB-laden sediment is a feasible and applicable
disposal action. However, the cost of incinerating PCB-laden sedi-
ment at a presently operating  incinerator is high due to the low
BTU value of the sediment and the high  supplemental energy re-
quirements to incinerate  the  sediment as required  by  40 CFR
761.40. This greatly exceeded the cost of sediment disposal at an
off-site chemical waste landfill, without providing a commensurate
environmental benefit. Since the high cost for incineration is based
on the energy requirements for  PCB destruction,  an on-site in-
cinerator would have a comparable cost. Therefore, neither on- nor
off-site incineration were considered further.

Disposal
  After the removal  of the contaminated sediment is complete,
solids dewatering for  transportation and disposal is required. The
disposal of wastes containing  concentrations of PCBs above 50
ppm  is regulated by TSCA while materials with PCB concentra-
tions below 50 ppm may be disposed of  under other federal and
state regulations.
  Based on the sampling data, the sediment removal volumes re-
quired to meet the 50,10 and 1  ppm concentration limits within the
river are 243,890, 472,405 and 1,286,760 yd3 respectively (Table 2).
However, the physical separation of sediment at various concentra-
tion levels is  not feasible. Also, the co-mingling of sediments to
achieve an average concentration of below 50 ppm (dilution) is pro-
hibited under TSCA.  In order  to remove  the 10 and 1 ppm levels,
the sediment with PCB  concentrations above 50 ppm must be in-
cluded. The distinction of sediment with  less than or greater than
50 ppm would require extensive sampling  and analysis prior to and
during removal.  Therefore, all sediment removed  would be re-
quired to be disposed  of at a licensed chemical waste facility.
                                                                       RISK ASSESSMENT/DECISION ANALYSIS
                                                         281

-------
  Assuming that a SO ppm limit was established, the transportation
of 243,890 yd3 of sediment would be required. At 12 yd3 per load,
this amounts to a total of 20,325 truckloads. Using six months per
year due to  weather conditions and dewatering requirements, ap-
proximately 1,130 trucks per month would travel  from the site(s)
via local, secondary and interstate roadways.
  Extensive  mitigating measures would be required to reduce the
many environmental  risks associated with this alternative.  The
potential for leaks and accidental spills with such a  large volume of
material and traffic is inherent.  The social impact of the transporta-
tion  of contaminated materials through residential areas is great
and has been a major determining factor at many other waste sites.

                           Table 2
             PCB Removal from Plant to State Border
PCB Con-
centration
Limit (ppm)
50
10
1

PCB
Removal
Ob)
25,533
31,617
36,297
«PCBi
Removed

69
86
98
Sediment Removal at
Add'tl *
PCBi
Removed
...
17
29
Sediment
Removal
(yd-)
243,890
472,405
1,286,760
Increased
Removal
Ratio-
—
2
5
"X" Concentration Limit
•Obtained by. 	
                      Sediment Removal at SO ppm
                            Table 3
                 Sediment Removal Requirements
ALTERNATIVE
50 PPM
10 PPM
1PPM
SEDIMENT VOLUME (YD3)
243.890
472,405
1,286,780
cosr
$18,291,750
$35,430,375"
$96,507,000"
 •Cost-J75/cubk yard assuming local disposal.
 •Thae com do not reflect the added coiu of transportation, mobilization, and establishment ot
 additional disposal sites beyond those required for JO ppm.
  The additional cost of transportation and disposal at an off-site
facility was not justified given that the same measures of protec-
tion and environmental benefit can be achieved by local disposal.
Therefore, the availability of local disposal sites will  be investi-
gated during the detailed engineering evaluation. These  sites are
required to meet the specification of 40 CFR. 761.41.
Extent of Remedy
  As  previously stated, the Consent Order required the  study of
remedial actions necessary to meet PCB concentration limits of 50
and 10 ppm. In addition, the company agreed to study the feasi-
bility  of the removal or treatment of sediments to PCB concentra-
tions of 1 ppm. However, the criteria which should be used to eval-
uate the remedial alternatives necessary to achieve these levels were
not specified. The NCP guidelines were again used to determine
                                                        the appropriate "extent of remedy" required during this privately
                                                        financed cleanup activity.  Specifically, the remedial alternative
                                                        chosen must be cost-effective and effectively mitigate and mini-
                                                        mize damage to and provide adequate protection of public health,
                                                        welfare and the environment.
                                                          The estimates of PCB load and sediment removal requirement*
                                                        for the river from the facility to the state line are presented in Table
                                                        1.  Sediment removal volumes and associated PCB removal in Ibs.
                                                        for the 50, 10 and 1 ppm concentration limits are included. Using
                                                        this information, the percentage of PCBs removed from the river
                                                        for each limit were developed and presented in Table 2.
                                                          To achieve the 50 ppm limit, a total of 25,533 Ibs. of PCBs would
                                                        be removed; representing 69% of the  total PCBs present in the
                                                        river. In addition, this requires the removal of 243,890 yd* of sedi-
                                                        ment. Only selected sediment removal from the  facility to and in-
                                                        cluding the pond would be required.
                                                          In comparison, to achieve the 10 ppm  limit, a total of 31,617 Ibs.
                                                        of PCBs had to be removed from the facility to the state border.
                                                        This represented only an additional 17% of the  total  PCBs re-
                                                        moved over the 50 ppm limit. In conjunction, this would require
                                                        twice the volume of sediment removal or 472,405 yd1.
                                                          The 1 ppm limit would result in the  removal  of 36,297 Ibs. of
                                                        PCBs or 1,286,780 yd* Ibs. of sediment from the facility to the state
                                                        border.  In comparison to the 50 ppm  limit,  an additional PCB
                                                        removal of only 29% would be achieved, requiring five times the
                                                        sediment removal volume required at 50 ppm. These figures rep-
                                                        resent sediment removal volumes only. The cost-effectiveness asso-
                                                        ciated with these removal volumes was  also analyzed.
                                                          Assuming a local disposal facility, the cost to remove and dispose
                                                        of PCB-laden sediment has been approximated to range from $50-
                                                        SlOO/yd'. Using an average of $75/yd',  the disposal cost would be
                                                        $18,291,750, $35,430,375 and $96,507,000 to meet the 50,10 and 1
                                                        ppm limits respectively (Table 3).
                                                          Based on the foregoing analysis, it was determined that the re-
                                                        moval of the total sediment volumes required to meet the 50, 10
                                                        and 1 ppm limits was neither reasonable nor cost-effective. This did
                                                        not preclude, however, the possibility  of limited  removal from
                                                        selected areas of sediment with PCB concentrations greater than 50
                                                        ppm, where other remedial measures are  ineffective in mitigating
                                                        environmental and public health impacts. Therefore, removal of
                                                        total sediment volumes required to meet the 10  and 1 ppm limits
                                                        were not evaluated further.
                                                          The NCP criteria were used as a guideline to determine the extent
                                                        of remedy required for PCB removal along the river as the 50 ppm
                                                        PCB concentration level was determined to be the most appropriate
                                                        level of cleanup in this case.
                                                          Government approval has been given to the above evaluation.
                                                        Currently, remedial action planning for only those areas with PCBs
                                                        greater than 50 ppm is being conducted.
                                                        CONCLUSIONS

                                                          The two case studies presented in this paper outline how the cri-
                                                        teria set forth in the NCP provided guidance during the remedial
                                                        action planning at several hazardous waste sites. In each case, the
                                                        settlement documents agreed upon by the responsible parties pro-
                                                        vided little decision-making criteria from which remedial alterna-
                                                        tives could be assessed  nor did they address the question, "How
                                                        clean is clean?" The NCP provided the specific guidelines used to
                                                        define the appropriate extent of remedy required at  these sites as
                                                        well as to evaluate, screen and select the most cost-effective and en-
                                                        vironmentally sound remedial alternatives.
282
RISK ASSESSMENT/DECISION ANALYSIS

-------
   EFFECTS  OF  UNCERTAINTIES  OF  DATA  COLLECTION
                                   ON  RISK  ASSESSMENT

                                         GARY L. McKOWN, Ph.D.
                                             RONALD  SCHALLA
                                            C. JOSEPH ENGLISH
                                   Battelle Project Management Division
                                  Office of Hazardous Waste Management
                                             Richland, Washington
INTRODUCTION
  The primary purpose of site characterization is to provide a data
base for use in determining whether a problem exists, the nature
and extent of the problem vis-a-vis potential remedial actions and
the resulting risk contaminants on the site pose to human health
or environmental systems. The primary means  for evaluating the
current situation is to compare  concentrations of chemicals in
samples obtained from within environmental pathways, measured
at or extrapolated to the point where receptors would be impacted,
with human health risk levels  and/or established environmental
standards. Evaluation of both risk and feasible remedial actions,
however, should also consider past and future  potential impacts.
Such considerations require data for understanding contaminant
sources and the mechanisms by which chemicals are, or would be,
transported from a source to a receptor.
  The uncertainty  in data acquired at a hazardous waste site is
often neglected in  assessment of observed or potential risks. For
the simple case of evaluating risks posed by exposure of receptors
to chemicals, the risk analysis will be limited by the degree of con-
fidence in contaminant levels as determined by sampling and analy-
sis. For determining mechanisms of source release and  transport
and for  evaluation  of  potential impacts and remedial action
efficiencies, uncertainties in data on source release, migration and
contaminant levels  at receptor sites all affect the assessment.
  In this paper, the authors discuss the uncertainties inherent in
geohydrologic and  chemical analytical field data, with consequent
effects on risk analysis  and decision making.  The uncertainties
and errors that can be introduced by  computations,  statistical
analysis or mathematical simulation have  been ignored; we have
assumed  that the  computation and interpretation methods are
conservative, and  any errors in  data will carry  forward to the
assessment.

UNCERTAINTIES IN GEOHYDROLOGIC DATA
  Because groundwater  impact is a major concern at most haz-
ardous waste sites,  understanding flow in groundwater systems is a
prerequisite for predicting contaminant pathways and rates of mi-
gration. Although  flow patterns can be determined using theoret-
ical methods  (e.g., analytical, numerical or  physical models),
empirical methods  based on laboratory and field measurements of
hydrologic parameters are the most commonly used approach in
hazardous waste migration  studies. The direct approach is most
appropriate because of the scale and precision required for the lim-
ited areal extent typical of the study sites.
  Measured parameters can be grouped into spatially variable par-
ameters and  parameters that are both spatially  and temporally
variable. Hydrologic parameters that are spatially dependent in-
clude:
•Horizontal hydraulic conductivity or transmissivity
•Vertical hydraulic conductivity or leakage
•Coefficient of storage or storativity
•Porosity, specifically effective porosity
•Specific yield and retention
  Those variables that are also temporally dependent include:
•Stress (recharge and discharge)
•Potential or hydraulic head
  The latter  three spatially-dependent parameters generally have
the smallest ranges and have more easily measurable and predict-
able properties than hydraulic conductivity. Hence, they have the
least uncertainty. For example, in unconfined aquifers of porous
media, values for storativity typically range from 0.05 to 0.25. In
unconfined aquifers, storativity is the same as specific yeild, which
is a parameter easily estimated from published values.1-2 The spe-
cific yield is approximately equal to the porosity or effective poros-
ity for more permeable materials. In any case, substituting specific
yield values is a useful and inexpensive method for estimating stor-
ativity and effective  porosity from geologic samples  or lithologic
descriptions. Because the spatial variability of these  three hydro-
logic parameters is relatively small, the potential undertainty is
quite small.  In  fact, an experienced  hydrogeologist can usually
select a representative set of values based on lithologic  data alone
that rarely is in error greater than a factor of one.
  While these parameters may be reliably  determined for uncon-
fined porous media, obtaining accurate values in confined systems
and fractured media is more difficult. Storativity values for  con-
fined aquifers may range from 0.00001 to 0.001, and the range
would be even greater for fractured media.3 Values  for effective
porosity and specific yield are still easily estimated for confined
porous media; these parameters are most difficult to predict in frac-
tured rock, particularly  if transmissivities are low.4 Collection of
representative data is difficult and costly, and the applicability of
data collection methods is limited.' Even when field procedures and
measurements are precisely controlled and accurate within ±2%,
the actual value obtained may have  large uncertainties.4- 6  Des-
pite these difficulties, uncertainties in the values obtained are small
compared to uncertainties in hydraulic conductivity.
  Hydraulic  conductivity or transmissivity determinations almost
always require direct measurements to confirm  preliminary esti-
mates or indirect measurements.7' 8 Direct subsurface measure-
ments require an access port, i.e., a monitoring well. Thus, a mon-
itoring well must be designed to allow reliable and representative
measurement  of hydraulic conductivity, potential  and chemical
                                                                     RISK ASSESSMENT/DECISION ANALYSIS
                                                       283

-------
quality of groundwater. The importance of monitoring well design
and configuration on accuracy, resolution and representativeness
of data has been well-documented.9-I0- "•l2-l3 Aspects which affect
the uncertainty of data are discussed below.

Well Diameter

  Transmissivity and storativity values obtained from small-diam-
eter monitoring  wells are  less representative, and the data uncer-
tainties are greater than for measurements on larger diameter wells.
In a recent example, transmissivities obtained  from tests on 2 in.
diameter wells were shown to be  in error by up to two  orders of
magnitude. The higher chemical  and potential data density pos-
sible through  use  of small-diameter wells15' " must be balanced
against the need  for data on other properties affecting migration.

Screened Interval
  Depth and selection of length of the screened interval must con-
sider the probably pathways for  contaminants to enter into and
move through the saturated zone.  Vertical placement and length of
the monitored  interval  must consider  both  hydrostratigraphic
equivalency and  chemical properties of the contaminants.''

Materials of Construction

  Materials used for casing and screen have only minor effects on
uncertainties in hydrologic parameters" but may have a large effect
on chemical data. Leaching of constituents from the well materials,
or adsorption of contaminants onto the materials, can occur.

Installation Techniques

  Improper drilling and installation methods can drastically affect
the measured  transmissivity, resulting in  errors of up  to a few
orders of magnitude."- "• x The most common installation prob-
lems are alterations of  the formation via drilling methods (mud
rotary drilling was recently shown to seriously affect  subsequent
transmissivity  measurements) and inadvertent hydraulic  intercon-
nection  between distinct hydrostratigraphic units. Also,  failure to
adequately develop a well may result in transmissivities with a high
degree of uncertainty.17-2l

Hydraulic Potential

  Hydraulic potential is obtained by direct field measurement, and
uncertainties can result from inaccurate measurements  of water
level and ground surface elevation." The principal uncertainty in
potential data, however, relates to the degree of hydrostratigraphic
and time equivalence of the measurements. Uncertainties in inter-
pretation of potential data can be equal  to or greater than  the hy-
draulic gradient, leading to gross  misinterpretation of flow direc-
tion.

Transmissivity

  Transmissivity or hydraulic conductivity can be determined by
laboratory methods. Olson and Daniel1' compared field and labor-
atory hydraulic  conductivity (K)  values from the literature and
found that field  results are usually higher than laboratory results
for the following reasons:
•Laboratory tests were usually performed on more clayey samples
•The presence of sand seams,  fissures and other macrostructural
 features are not represented properly in laboratory tests
•Laboratory K values back-calculated from consolidation theory
 were often used instead of directly measured values
•Vertical flow K is usually measured in the laboratory, whereas
 horizontal flow generally occurs in the field
•Distilled water is normally used in the laboratory tests
•Laboratory samples often have air entrapped in the sample
  Other  sources of error in  laboratory testing include: voids
formed during sample preparation, smear zones, growth  of micro-
organisms, use of excessive  hydraulic gradients  and temperature
effects.  A  large uncertainty in hydraulic conductivity  or trans-
missivity can be expected if reliance is made on laboratory-derived
values.
   While field testing to measure hydraulic conductivity or trans-
missivity may eliminate many of the problems associated with lab-
oratory tests, other potential errors can be caused by:
•Unrepresentative values caused by  inappropriate drilling  tech-
  niques or well construction methods as discussed above
•Inadequacy of the test method to provide data on heterogeneity,
  anisotropy or storativity of the aquifer and confining units
•Use of simplistic conversion factors in calculating transmissivity
•Use of a formula to calculate transmissivity or hydraulic conduc-
  tivity which is inappropriate for the actual  field conditions
   Any deviation in field conditions from the theoretical conditions
assumed in derivation of a formula will lead to some amount of
error in the computed values.  These deviations should be taken
into account when final evaluation of the  field test data is made.
In practice, the  formulas have been applied with success and the
calculated hydraulic characteristics  have proven to be reliable for
most purposes." Uncertainties in field data of the type discussed
above, however, will be transmitted through any calculations.
UNCERTAINTIES IN CHEMICAL DATA

  The uncertainties  in  the  concentration of  various  chemical
species at or neat sources, within environmental  pathways or at
points of exposure depend on two factors:
•The uncertainties  in collection of representative samples from
 sources, environmental pathways or exposure points and presen-
 tation of the samples for chemical analysis
•The uncertainties in determination of the constituency of samples
 that will define health or environmental risks
Sampling Methodology Uncertainties

  Sampling involves the acquisition of a small piece of the environ-
ment that is representative of the entire matrix to be considered in
risk  analysis  in such a manner  that the representativeness is not
compromised. When performed in site investigation programs with
limited budgets and schedules, sampling is beset  with uncertain-
ties both in theory and in practice. Statistical analysis may require
many replications of data before confidence limits become accept-
ably small. Methods accounting for data variability require an even
larger data set over the  parameter space. In the  authors'  exper-
ience, neither time nor funds are usually available for the extensive
sampling efforts needed to establish a high degree of confidence in,
and  tight confidence bounds on, environmental data. A single
sample is used more often than replicates as the basis for a risk
analysis, and the very low confidence that results is usually ignored.
  A particular case where sampling theory is at  odds with budget
and  schedule involves neglect of factors known to affect the valid-
ity of risk analyses but requiring extensive field data for complete
definition. In a recent example, careful and replicate procedures
were employed to precisely define the chemical profile within a
soil  column.  Following  this determination, the results were cor-
rected  by  an assumed  soil-water partition  coefficient (perhaps
valid to within I -2 orders of magnitude) for comparison with an
established health risk water criterion. Although  the uncertainty in
the resulting  assessment  was recognized, no attempt was made to
validate the assumption,  because the soil partitioning experiments
would  have forced a delay in site assessment of at  least several
weeks. In  effect, the exacting soil column sampling was wasted
effort and did not enhance the certainty of the assessment.
  In any sampling design, the relative uncertainty of all factors in a
risk  analysis needs to be  considered because the uncertainty in the
analysis can be no less than that of the  least certain factor. Avail-
able time and funds would be better spent in determining factors
with the greatest degree of uncertainty and with the greatest impact
on the overall uncertainty of the assessment.
284
          RISK ASSESSMENT/DECISION ANALYSIS

-------
Sampling Procedure Uncertainties

  Uncertainties and biases in the sampling procedures can have a
significant effect on data assessment.  Some  obvious factors,  as
well as concepts which may not be obvious to site investigators,
are considered  below.  Because site characterization  efforts  gen-
erally involve collection of groundwater samples, the discussion
focuses on this type of sample.
  Collection of groundwater samples from an appropriately-de-
signed monitoring  well  network  would seem to  be relatively
straightforward; procedures are generally specified in  most site in-
vestigation programs. However, there are a number  of instances
where generally acceptable procedures can be inadequate or can re-
sult in gross uncertainties in assessment data. Several cases that are
frequently encountered during site investigations are considered be-
low. The possible uncertainties for two of  the cases (inadequate
purging of wells and cross-contamination) are considered in detail.

Inadequate Well Development and Purging

  Drilling and completion of monitoring wells represents  extreme
disturbance of the natural groundwater matrix. Development  or
purging of wells is essential to remove foreign materials introduced
during the drilling arid to restore the groundwater to reasonably
representative  conditions.  Similarly,  wells  should  be   purged
immediately prior to sampling  to remove water in the well  bore
and the formation immediately surrounding the  well bore.  In
either case, approximately 5 equivalent volumes (EV)  are typically
removed, and even this quantity is  often not removed for  wells  of
very low yield.
  To consider  the  uncertainties introduced  by inadequate  well
development  or  purging, we have calculated possible  residual
effects after removal of 5 EV from  a typical well of 8  in. borehole
diameter,  4 in.  casing diameter, 15 ft sand  pack (30% porosity),
10  ft screened interval and 30  ft water column. The volume  of
standing water (1 EV) is 3.8 ft3 (107.5 L) and 5 EV is 538 L. For
contaminants introduced by the drilling process, or for variations
in chemical composition between the open well and the formation,
a dilution factor of 538 can be expected between the volume purged
and the typical 1 1 sample collected  for analysis. If the action  level
for a contaminant is 1 jig/1 (criteria for many contaminants are
even lower), the criterion in the sample  would be exceeded if  only
538 jig of contaminant were introduced during the drilling or sam-
pling process. Drilling would involve removal of about 3 x  10! g  of
formation material from just the sand-packed region, and sampling
would involve purging of about 5 x 105 g of water. Based strictly on
a materials-handling concept, the odds against avoiding introduc-
tion of  538 ug of "foreign" material are 6  x 10' and 9 x 10" for
drilling and sampling, respectively.  In practical terms, the removal
of 5 EV during development  or sampling could be totally inade-
quate for providing data with a high degree of certainty.
  Similar considerations apply, in reverse, if clean water is intro-
duced during drilling into a highly contaminated aquifer or if sam-
pling  for volatiles (which escape from the water in an open well) is
attempted. For these cases, the measured concentrations following
purging will be lower than those of the natural matrix.
  It is apparent that any reasonable degree of well development  or
purging will not assure removal of foreign materials introduced
during drilling or sampling. The authors recommend a number  of
procedures to minimize the problem:
Purging
  Greater dependence should be placed on natural purging of mon-
itoring wells. A considerable period of time should transpire be-
tween well completion and sampling. For most aquifer systems, 2
to 4 weeks prior to sampling is inadequate. If the groundwater flow
rate can be estimated, then the time required to exchange water
within the well  bore can be calculated. For the typical well con-
sidered above, and a typical flow rate of 0.5 ft/day,  the time re-
quired to purge 100 EV via the natural groundwater flow would
be 76 days. If the schedule permits, a waiting period of 2 to 3
months would be recommended before sampling the well. An ad-
vantage of natural purging is that no hazardous wastes are gen-
erated, as is often the case for water removed from contaminated
wells.

Sample Series
  A series of samples should  be obtained from the well  over a
period of time. The presence of induced contamination can be de-
rived from the change in analytical data with time. The possibility
that natural changes in groundwater contaminant levels are occur-
ring, however, may require collection of several samples so that
the contribution arising from induced contamination can  be cal-
culated.
Induced Contamination
  The possibility that low levels of contaminants could be due to
induced contamination should always be considered in risk assess-
ments. Given an understanding of the source location and the hy-
drologic system, the uncertainty in the risk analysis from this factor
can be calculated. In general, the degree of uncertainty will be high
for a small number of samples acquired shortly after well comple-
tion and limited development.

Cross Contamination During Sampling
  The potential for inducing foreign materials into a well via sam-
pling equipment should always be considered. Sampling of any well
after sampling of highly contaminated wells increases the potential
for cross contamination. At one hazardous waste site investigated
in 1983, the authors observed contamination in wells ranging from
essentially pure methylene chloride (1,330 g/1) down to the detec-
tion limit of about 2 ug/1. The amount of pure methylene chlor-
ide that would need to be transferred to the typical well to produce
2 Jig/1 after purging of 5 EV (538 L) is 8 x 10~7 1,  or about one-
tenth of a drop.
  Considering the potential for cross contamination and the diffi-
culties involved with stringent decontamination of sampling equip-
ment in the field, the authors recommend the following procedures:
•Individual sampling equipment, including  bailers  and pumps,
 should be used in each well. No sampling equipment should be
 transferred between wells.
•Field operational planning should specify sampling of wells in
 order of increasing expected concentrations of contaminants. For
 highly contaminated wells, sampling equipment should  be dis-
 carded after use and not reused.
•The uncertainty arising from potential cross contamination should
 be factored into any subsequent risk assessment.

Other Elements of Uncertainty
  A number of other potential problems associated with sam-
pling and analysis must be considered in evaluating the uncertainty
of risk assessments:
  If fracture flow and circuitous  flow paths of groundwater are
suspected, analytical data from any given well may be highly uncer-
tain for assessing contaminant migration. For such cases, the chem-
ical data may make no sense until the major flow paths have been
defined.
  To evaluate the quality of the entire sampling  and analysis pro-
tocol, QC samples should be introduced into sampling lots in  the
field. However, this practice is difficult to implement with a high
degree of certainty. The authors have observed benzene at 33 jig/1
in distilled water used for field blanks.
  Sampling methods employed require consideration of the types
of analytes; the use of most pumps and bailers can result in loss of
volatiles, and aeration of samples may enhance aerobic chemical or
biological transformations. A  detectable loss  of  volatiles and
phenols from  samples has been observed after using even low-
energy bladder pumps in monitoring wells.
  Filtration of samples can result in large uncertainties; the poten-
tial for cross contamination is increased, and contaminants can be
adsorbed onto the filter body and filtration medium. The  authors
have observed considerable loss of phenols from water samples
during filtration, due to absorption onto filters.
                                                                        RISK ASSESSMENT/DECISION ANALYSIS
                                                         285

-------
   Preservation, shipping and storage of samples prior to analysis
 are important considerations. Possible effects include chemical and
 biological degradation as a result of the unnatural state of bottled
 samples and the potential for adsorption of contaminants on con-
 tainer wells. The authors have observed degradation, and consid-
 erable  adsorption onto glassware, of phenols  within a  few  hours
 after sampling.
   The very small quantity of contaminants required to produce de-
 tectable concentrations can be introduced  into water samples via
 air transport. The authors have observed detectable levels of aro-
 matic hydrocarbons in wells that were  traceable to  use of a  pump
 driven by a gasoline generator.
   A major problem with soil,  sediment or waste  samples  is the
 difficulty of achieving representativeness. Even when extreme pre-
 cautions are taken during collection, grinding,  sieving and subsam-
 pling,  large uncertainties can be avoided only  by analyzing a large
 number of samples.
   Cross contamination during solids sampling is highly probable
 unless stringent equipment  decontamination  and  elaborate pro-
 cedures are employed. A high  degree  of uncertainty  is expected
 for samples collected from a soil boring because of the potential for
 cross contamination.
   Analysis should be performed on solid samples in the wet state.
 The practice of air drying samples before analysis results in loss of
 volatile and semivolatile compounds and  increases the potential
 problems with chemical and biological degradation many-fold.
   Despite elaborate  procedures and extreme  care  in conduct of
 sampling and analysis, considerable uncertainty in analytical data
 will  remain. Most analytical methods will provide data that are, at
 best, good only to within  ± 10-20%. The validated analytical un-
 certainty  should be carried forward as a "best case" for the risk
 assessment and remedial action analysis.
  CONCLUSIONS
    The considerable effects that arise from uncertainties in field
  data should be considered in any subsequent assessments and de-
  cisions. The potential for uncertain results should be quantified
  and considered in planning of field efforts. The manner in which
  field work would be conducted to minimize uncertainties in subse-
  quent assessments  is at  odds with typical  schedules and budgets
  for site investigations.
  REFERENCES

   1. Johnson,  A.I.,  "Specific Yield—Compilation of Specific Yields for
     Various Materials,"  U.S.  Geological  Survey  Water Supply Paper
     1662-D, 1967.
   2. Evenson,  R.E.. Wilson, H.D., Jr. and Muir, K.S., "Yield of the
     Carpenteria and Goleta Ground Water Basins, Santa Barbara Coun-
     ty, California, 1941-58," U.S. Geological Survey Water Supply Paper
     1470,1959.
   3. Johnson,  E.E., Inc., Ground water and  Wells,  Johnson Division,
     UOP Inc., Saint Paul, MM, 1975.
   4. Gonzales, D.D. and Bentley, H.W., Field  Test for Affective Porosity
     and Dispersivity in Fractured Dolomite, the WIPP, Southeastern New
     Mexico, AGU Water Resources Monograph 9, 1984, 207-221.
   5. Snow, D.T., "Rock Fractures, Spacing!, Openings, and Porosities."
     J. of Soil Mechanics, Foundations Division, ASCE, 1968, 73-91.
                                                             6. Grove, D.B. and Beet em, W.A., "Porosity and Dispersion Constatat
                                                               Calculations for a Fractured Carbonate Aquifer Using the Two Well
                                                               Tracer Method," Water Resources Research. 7, 1971,128-134.
                                                             7. Heigold. P.C., Gilkeson,  R.H.,  Cartwright,  K. and  Reed,  P,C.,
                                                               "Aquifer Transmissivily from Surficial Electrical Methods," Ground
                                                               Water, 17. 1979,338-345.
                                                             8. Glaccum, R.A., Benson, R.C. and Noel, M.R., "Improving Accuracy
                                                               and  Cost-Effectiveness of Hazardous Waste Site  Investigations."
                                                               Ground Water Monitoring Review. 2. No. 3, 1982. 36-40.
                                                             9. Barvenik, M.J. and Cadwgan, R.M., "Multilevel Gas-Drive Sampling
                                                               of Deep Fractured Rock Aquifers in Virginia," Ground Water Mon-
                                                               itoring Review. 3. No. 4,1983, 34^0.
                                                            10. Ward, J.R., "Considerations in Multi-Aquifer Monitoring for In Situ
                                                               Leach Mining," Ground Water Monitoring Review, 3, No. 1,  1983,
                                                               118-121.
                                                            II. Johnson, T.L., "A Comparison of Well Nests vs. Single-Well Com-
                                                               pletions," Ground Water Monitoring Review. 3. No. 1, 1983,76-78.
                                                            12. Desaulniers, D.E., "Hydraulically Driven Piezometers for Monitoring
                                                               in Soft Sediments," Ground Water Monitoring Review. 3, No. 2,
                                                               1983.16-20.
                                                            13. Cherry, J.A. and Johnson,  P.E., "A Multilevel Device for Monitor-
                                                               ing in Fractured Rock," Ground  Water Monitoring Review, 2, No. 3,
                                                               1982,41-44.
                                                            14. Voytek, J.E., "Considerations in the Design and Installation of Mon-
                                                               itoring Wells." Ground Water Monitoring Review, 3, No.  I.  1983,
                                                               70-71.
                                                            IS. Devary, J.L. and Schalla, R.,  "Improved Methods of Flow System
                                                               Characterization," Proc. National Conference on Management of Un-
                                                               controlled Hazardous Waste Sites, Washington,  D.C.,  Nov., 1983,
                                                               117-122.
                                                             6. Palumbo, M.R. and Khaleel. R., "Kriged  Estimates of Transmissiv-
                                                               ity in the Mesilla Bolson, New Mexico,"  Water Resources Bulletin,
                                                               19. 6, 1983,929-936.
                                                            17. Schalla. R., McKown, G.L., Meuser. J.M., Parkhurst, R.G., Smith,
                                                               C.M., Bond, F.W. and English, C.J., Source Identification,  Con-
                                                               taminant Transport Simulation, and Remedial Action Analysis, Anna-
                                                               ton Army Depot. Anniston, AL.  DRXTH-AS-CR-83265, U.S. Army
                                                               Toxic and Hazardous Materials Agency, Aberdeen Proving Ground,
                                                               MD, 1984.
                                                            18. Clark, L. and Turner,  P.A.. "Experiments to Assess the Hydraulic
                                                               Efficiency of Well Screens." Ground Water. 21. 1983.270-281.
                                                            19. Fetter, Jr.,  C.W., "Potential Sources of Contamination in Ground-
                                                               water Monitoring,"  Ground  Water Monitoring Review, 3,  No. 2,
                                                               1983,60-64.
                                                            20. Brobst, R.B.  and Buszka,  P.M..  "Effects of Mud  Rotary Drilling
                                                               Methods on  Hydrogeological  Information from Monitoring Wefls,"
                                                               Proc. of the 4th National  Symposium and Exposition on Aquifer
                                                               Restoration and Ground Water Monitoring, 1984.
                                                            21. Clark. L.. "Development of Wells is Vital." The Johnson Drillers
                                                               Journal, 55, Nos. 3 and 4, 1983, 12-14.
                                                            22. Schalla, R. and Leonhart, L.A.. "Dealing with Regional Hydrologk
                                                               Data Base Limitations and Uncertainties, Case Example: The Colum-
                                                               bia River Basalts," Proc. of the Symposium of the Effectiveness of
                                                               Geologic  Isolation  of  High Level Radioactive  Waste, Gatlinburg,
                                                               TN. 1981,415-423.
                                                            23. Olsen, R.E. and Daniel, D.E., "Measurement of Hydraulic Con-
                                                               ductivity  of Fine-Grained  Soils," In:  Permeability and Ground-
                                                               Water Transport, ASTM STP 746, American Society for Testing and
                                                               Materials. Philadelphia. PA, 1981,18-64.
                                                            24. Kruseman, G.P. and De Ridder, N.A., Analysis and Evaluation of
                                                               Pumping Test Data, Bull. No. II, International Institute for Land
                                                               Reclamation and Improvement, Wageningen, Netherlands, 1976.
286
RISK ASSESSMENT/DECISION ANALYSIS

-------
AN EPIDEMIOLOGIC  STUDY  OF COMMUNITY EXPOSURES
             TO2,3,7,8-TETRACHLORODIBENZO-P-DIOXIN
                                               PAUL A. STEHR
                                                GARY F. STEIN
                                                 KAREN WEBB
                                       Center for Environmental Health
                                 Department of Health and Human Services
                                                Atlanta, Georgia
CHARACTERIZATION OF ENVIRONMENTAL
CONTAMINATIONS
  In 1971, approximately 29 kg of 2, 3, 7, 8-tetrachlorodibenzo-
p-dioxin (TCDD)-contaminated sludge wastes, which originated as
a byproduct of  hexachlorophene production in a  southwest
Missouri plant, were mixed  with waste oils and sprayed for dust
control in that part of the state. Two hundred forty-one residential,
work and recreational areas (including  several horse arenas), in-
cluding the town of Times  Beach, were thought to be contam-
inated. As of January, 1984, 36 sites have been confirmed as hav-
ing  at least 1 ppb of TCDD in soil, 98 have not shown contam-
ination at this level and an additional 107 are still under investiga-
tion. Levels as high as  35,000 ppb were originally measured in soil
at one of these 36 sites; current isolated levels as high as 1,900 ppb
exist in these contaminated areas, but most detectable levels in soil
samples range from less than one ppb to several hundred ppb.
  About half of the 36 confirmed sites are contaminated with peak
levels in excess of 100 ppb; 11 of these (69%) are in residential
areas. These sites vary widely in  their potential for leading to
human exposure  due  to the lack of uniformity in  geography,
topography, geology and characteristic  land use. This variability
has presented difficulties in the public health policy decision-mak-
ing  process. Sites which have high levels of contamination and are
in areas of frequent and regular access constitute the greatest public
health risk. At other  sites,  however, dioxin contamination is in
clearly circumscribed areas,  at subsurface depths exceeding 15 ft,
under paved areas or in areas with limited land use. All of these
considerations were taken into account in assessing the risk  of ex-
posure for an estimated 4,600 individuals from these contaminated
areas from 1971-1983.
  The earlier phases of this investigation focused on several sites in
eastern Missouri; later activities included all 36 contaminated sites.
The Centers for Disease Control (CDC) had previously worked
with the Missouri Division of Health (MDH) in 1971 at the time the
initial contaminations  occurred after receiving a report of an ex-
posed child who presented with hemorrhagic cystitis; in 1974, this
work culminated in the laboratory identification of TCDD  in the
waste oil. With further discoveries of widespread contaminations
in mid-1982, MDH and CDC in consultation reinitiated public
health activities on the basis of new information and additional en-
vironmental data.
RISK TO HUMAN HEALTH
  The case of dioxin illustrates many of the difficulties encountered
in assessing health risks following long-term, low-dose exposure to
environmental chemical contaminations. As yet, there is no re-
liable, widely available method for directly measuring dioxin levels
in humans. In this investigation, the lack of any direct measure of
body burden substantially hindered attempts to assess the degree of
exposure to and concomitant health risk posed  by environmental
dioxins.
Exposure Assessment

  Therefore, the long-term risk of exposure in  any area contam-
inated with dioxins must be determined by considering the excess
risks of developing specific adverse health effects as a result of an
estimated total cumulative dose. This dose is a function of several
factors: (1) the  concentration of environmental contamination,
(2) location of and access to contaminated areas, (3) the types of
activities conducted in contaminated areas and (4) duration of ex-
posure. These assessments were concerned primarily with health
risks in regard to contamination of soils in residential areas.
  To estimate exposure, the authors made assumptions regarding
the bioavailability and absorption of TCDD from soil as well as
other metabolic parameters. Moreover, principal routes of uptake
were thought to be through dermal absorption, ingestion and in-
halation of contaminated dirt/dust particles.
Risk Assessment

  Animal studies have shown great species variability in both acute
and chronic responses to TCDD exposures; where humans fit on
this response scale  is not clear. However, common findings from
both animal toxicological work and limited data on cases of high
dose, exposures  of humans have indicated prominent effects on
several organ systems: liver changes include diminished function,
hepatocellular  necrosis,  tumor  induction  (in  animals)  and
microsomal enzyme induction; other effects include chloracne, de-
pressed cell-mediated immunity and peripheral neuropathy. Addi-
tionally, some studies have suggested that occupational exposures
to TCDD may induce an excess risk of developing soft tissue sar-
comas, but the only adequate dose-response data available for use
in the risk assessment calculations  were from animal carcinogen-
icity studies. A linear, nonthreshold dose-response model was used
to calculate increased lifetime cancer risk, and the calculation
methods incorporated guidelines that  a group of outside consul-
tants recommended to CDC.

Risk Management

  Based on these calculations, the authors concluded that residen-
tial soil TCDD levels of ZL 1 ppb pose a level of concern for de-
                                                                   RISK ASSESSMENT/DECISION ANALYSIS
                                                                                                                  287

-------
layed health risks. In highly contaminated areas (areas with soil
contamination levels ^ 100 ppb), with a high degree of access and
concomitant exposure, the estimated incremental lifetime cancer
risk may increase rapidly and be orders of magnitude higher than
1 per million. Therefore, MDH and  CDC issued advisories which
stated that the continued, long-term  exposure to persons living in
specified residential areas with 1 ppb or more TCDD contamina-
tion in the soil posed an unacceptable  health risk.
  These public health advisories and USEPA's consideration of the
available remedial options were the basis on which  site-specific
decisions to eliminate or  mitigate these exposures were made. The
time  frame for  such decisions was  dependent on the degree of
contamination and on the degree to which continued exposures
could be prevented while temporary or permanent remedial actions
were considered  and/or executed. In most cases, the USEPA opted
for temporary environmental cleanup, stabilization or restriction of
access to contaminated areas because of limited, well-defined areas
of contamination, relatively low TCDD soil levels or relative in-
accessibility of  contaminated areas. However, in  several  note-
worthy situations (such as the case of Times Beach), it  was decided
that permanent relocation of residents was the most prudent action.

PUBLIC HEALTH ACTIVITIES
   In addition to ongoing review and assessment of EPA environ-
mental  sampling data, MDH and CDC began  four distinct public
health actions in January, 1983:
•Providing health education for both the medical and public health
  community and the general public about current understandings
  of the health effects ofdioxin exposures. A summary of the med-
  ical/epidemiological literature  was prepared  and  sent to phy-
  sicians in eastern Missouri. On Jan. 18, 1983, experts from gov-
  ernment, academic institutions and industry were  brought to-
  gether to give  a  seminar for the local medical community. In-
  dividual consultations and toll-free hotlines were established to
  answer questions from and concerns of the general public.
•Providing a dermatologic screening clinic to the general public.
  This clinic was intended to screen for cases  of chloracne as an in-
  dication of possible dioxin exposure. In February,  1983, on con-
  secutive weekends, all  residents of eastern  Missouri who had
  reason to suspect that they had been exposed and who had current
  skin problems were invited to be seen at these screening clinics.
•Creating and maintaining a central  listing of potentially exposed
  individuals. This listing will enable public health agencies to keep
  in touch with and locate potentially exposed individuals for edu-
  cational  purposes  or  possible epidemiologic  and/or  clinical
  follow-up. Specifically, when  a reliable screening  method for
  TCDD in serum becomes available,  we will be better able to assess
  their exposure  status and concomitant health risks. Baseline and
  identifying information was collected  in the  form of a Health
  Effects Survey questionnaire designed to  elicit information on
  possible routes' of exposure, life-style habits, residential histories,
  occupational histories and medical histories. It was also intended
  to serve both as a screening tool for identifying a "highest risk"
  cohort on whom intensive  medical evaluations were  focused and
  as a method of compiling a community-based data set from which
  epidemiologic inferences might be drawn.
•Designing and  implementing a pilot medical study of a ' 'highest
  risk" cohort. This  research was conceived  as a pilot study of a
  group of persons presumed to be at highest risk of exposure to en-
  vironmental TCDD. It  was intended to provide preliminary in-
  formation on  possible  health  effects  from  these exposures to
  enable investigators to develop more  refined  and specific epi-
  demiologic protocols to be used in further investigations.


INVESTIGATIONAL METHODS

   In this study, the authors assessed potential health effects related
to dioxin exposures by three means. First, as previously mentioned,
a Health Effects Survey questionnaire was developed to elicit in-
formation on each  person's exposure risk, medical  history and
                                                        potentially confounding influences. The authors sought data for
                                                        individuals believed to  be at risk of exposure  because they lived
                                                        near, worked at or frequently participated in activities near a con-
                                                        taminated site.
                                                          Second, a dermatology screening clinic was held.
                                                          Third, the authors reviewed approximately 800 completed ques-
                                                        tionnaires and selected 122 persons for inclusion in a pilot medical
                                                        study. A selected high-risk group of 82 individuals reported:
                                                        •Living or working in TCDD-contaminated areas or
                                                        •Participating on an average of more than once per week in activ-
                                                          ities that  involved close contact with the soil (such as garden-
                                                          ing, field/court  sports, horseback  riding,  playing in soil)  in
                                                          TCDD-contaminated areas with TCDD levels of between 20 and
                                                          100 ppb  for at least 2  years or levels greater than 100 ppb  for at
                                                          least 6 months. A low-risk  comparison group of 40 persons re-
                                                          portedly having  had  no access to  or regular  high-soil-contact
                                                          activities in any known contaminated areas was also selected.
                                                          Of the 122 persons in the  study group,  17.1% of the high-risk
                                                        group and lO.O^o of the low-risk group either refused to partici-
                                                        pate or failed  to appear for the examinations (this difference was
                                                        not  significant at  the 0.05 level), yielding a study population of
                                                        104 (68 at high risk and 36 at low risk of exposure).
                                                          In  addition  to being compared according to their responses on
                                                        the Health Effects Survey questionnaire, these 104  persons  were
                                                        assessed under a clinical protocol that included the following ele-
                                                        ments:
                                                        •Physical examination
                                                        •Neurologic examination
                                                        •Dermatologic examination
                                                        •Laboratory analyses
                                                        •Immune Response Tests
                                                        •Serum for use in TCDD analyses when such  tests become  avail-
                                                          able

                                                        RESULTS
                                                          The high- and low-risk groups were comparable in terms of age,
                                                        race, sex, education of head of household and interview respon-
                                                        dent  distributions.  The two groups did not differ significantly in
                                                        reporting other potential sources of exposure or the use of prescrip-
                                                        tion medicines. In regard to potentially confounding factors (such
                                                        as employment  in hazardous occupations or service in Vietnam),
                                                        there were no differences. The only significant difference in life-
                                                        style habits was that the high-risk group reported exercising more
                                                        regularly (< 0.01).
                                                          The authors found no differences  or consistent trends regard-
                                                        ing the prevalence of specific generalized  disorders as reported in
                                                        the  questionnaires,  the results of the  general physical  examina-
                                                        tions or the routine hematology tests (except  for a higher  mean
                                                        platelet count  and a nonsignificant trend of diminished peripheral
                                                        pulses in the high-risk group).
                                                          No consistent overall trends or statistically significant individual
                                                        diagnostic  differences were  detected for reproductive health out-
                                                        comes from the questionnaire material. No birth defects were re-
                                                        ported among children born to women in the high-risk group after
                                                        the time at which exposures could have occurred.
                                                          In the dermatologic screening, no cases of chloracne were seen
                                                        in the 140 persons examined or in the 104 persons in the study pop-
                                                        ulation. In addition, no significant differences in dermatological
                                                        findings were demonstrated by either medical histories or physical
                                                        examination for the study population.
                                                          Results of the neurological examinations showed no significant
                                                        differences  or  patterns between the  two  groups for the self-re-
                                                        ported neurological conditions or from the neurological examina-
                                                        tions.
                                                          As reported  in the medical histories, there were no differences
                                                        in prevalence of immune disorders. On physical examination, the
                                                        only significant difference noted was a suggestion of a greater prev-
                                                        alence of palpable nodes in the low-risk group. Laboratory analyses
                                                        showed no differences  between the two groups in regard to total
288
RISK ASSESSMENT/DECISION ANALYSIS

-------
induration in response to the antigenic skin tests,  the in vitro
lymphocyte  proliferative   responses   or  in  comparisons   of
parameters from T cell subset assays.
  In regard to the hepatic system, no trends or significant specific
problems were reported in the medical histories. On physical ex-
amination, there was a greater prevalence of hepatomegaly in the
high-risk group, but this finding also was not statistically signifi-
cant. There were no statistically significant differences between the
two groups on tests of hepatic  function. More specifically, the two
groups showed no difference in urinary porphyrin patterns, and no
cases  of overt porphyria cutanea tarda (PCT) or any  precursor
conditions (latent PCT or Type B porphyria) were detected.
  There appeared to be a  trend  of increased urinary tract prob-
lems among  the high-risk cohort on the basis of the medical his-
tory section of the questionnaire, although no statistically signifi-
cant differences were demonstrated. Urinalyses also suggested a
consistent pattern of abnormal findings, with  a  non-statistically
significant higher prevalence  of leukocyturia (.> 5 WBC/hpf)
and microscopic hematuria (>3 RBC/hpf) in the high-risk'group.

DISCUSSION

  The analyses  did not produce any firm indications of increased
disease prevalence directly related to the putative exposures. These
results  do,  however,  offer some insights and leads  for further
study. Of interest is the apparent trend indicative of urinary tract
abnormalities in the high-risk group, especially in light of the prev-
iously reported finding of hemorrhagic cystitis in  an exposed per-
son. The findings of no significant differences in liver function are
important; however, hepatic function should  be examined in  sub-
sequent studies because of other data suggesting hepatotoxic effects
of TCDD.
  Although  none of the findings from the immune function tests
and assays demonstrated statistically significant differences, several
results such as the indication of an increased  prevalence of helper
were of note: suppressor T-cell ratios < 1.0 in the high-risk group,
although the functional tests  of the immune system revealed no
overall abnormalities. In light of the rapidly evolving work in this
area,  follow-up and/or further investigation of these effects in ex-
posed cohorts should be conducted before drawing conclusions.
  Several factors could explain at least part of the overall negative
findings:
•The power to  detect significant differences  was restricted by the
  relatively small sample size
•A large percentage of the pool of persons from which the study
  and comparison groups were  chosen was self-selected, thereby in-
  troducing potential biases
•Because of the absence of an objective direct measure of exposure
  status, the possibility of individual misclassification errors exists
•Inability to detect effects with  long latency  periods  or subtle
  health effects for which our tests were  not sensitive
•It is conceivable that uptake  of dioxin from contaminated soils
  was generally less than estimated for this study group
•Chronic exposures to environmental TCDD have actually induced
  little or no adverse health effects

CONCLUSIONS
  These actions represent  the first phase in the  investigation of
dioxin  contaminations in  Missouri. The involved public  health
agencies continue to review environmental sampling data on new
suspected sites and develop public health advisories. Although the
results appear to be largely negative, no overall definitive conclu-
sion should be  based just on the initial pilot study. More refined
epidemiologic studies employing different designs and strategies are
planned to test the results of this pilot study. Concurrently, re-
search into replicable laboratory methods for measuring TCDD
body burden or other direct indices of exposure will be pursued.
  Finally,  public health policy in situations such as this  environ-
mental contamination with TCDD must continue to be focused on
the prevention of any potential health effects, even if such effects
were not demonstrated in a small pilot  study. For this reason, all
appropriate efforts need to be made to prevent human exposure.

SELECTED REFERENCES

 1.  Allen, J.R., Barsotti, D.A.,  Lambreckt, L.K. and Van Miller, J.P.,
    "Reproductive effects of halogenated aromatic hydrocarbons on non-
    human primates," Ann. NYAcad. Sci. 320, 1979,419-425.
 2.  Bauer,  H.,  Schulz, K.H.  and Spiegelberg,  U., "Berufliche  Ver-
    giftungen bei der Herstellung von Chlorphenol-Verbindungen," Arch.
    Gewerbepathol Gewer Behyg 18, 1961, 538-555.
 3.  Bleiberg, J., Wallen, M., Brodkin, R. and Applebaum, I.L., "Indus-
    trially acquired porphyria, "Arch. Dermatol. 89,1964, 793-797.
 4.  Carter,  C.D.,  Kimbrough,   R.D.,  Liddle,  J.A.  et al., "Tetra-
    chlorodibenzodioxin: an accidental poisoning  episode  in   horse
    arenas," Science, 188, 1975, 738-740.
 5.  Eriksson,  M.,  Hardell,  L., Berg, N., Moller, T. and Axelson, O.,
    "Soft tissue, sarcomas and exposure to chemical substances: a case-
    referent study," Brit. J. Med. 38,1981,27-33.

 6.  Fillipini, G., Bordo, B., Crenna, P.,  Massetto, N., Musicco and
    Boeri,  R., "Relationship between  clinical and  electrophysiological
    findings and indicators  of heavy exposure to 2, 3, 7,  8-TCDD,"
    Scand. J. Work Environ. Health 7, 1981, 257-262.
 7.  Gupta, B.N., Vos, J.G., Moore, J.A., Zinkl, J.G. and Bullock, E.G.,
    "Pathologic effects of 2, 3, 7, 8-TCDD in laboratory animals," En-
    viron. Hlth. Per. 5, 1973, 125-140.
 8.  Hardell, L. and Sandstrom, A., "Case-control study: soft tissue sar-
    comas  and  exposure to phenoxyacetic acids and  chlorophenols,"
    Brit. J. Cancer 39, 1979, 711-717.
 9.  Hook, G.E.R., Haseman, J.K. and Lucier, G.W., "Induction and
    suppression  of hepatic and  extrahepatic microsomal foreign-com-
    pound-metabolizing  enzyme  systems by 2,  3, 7, 8-TCDD," Chem.
    Biol. Interactions 10, 1975, 199-214.
10.  Hook, J.B., McCormack, K.M.  and Kluwe, W.M., "Renal effects of
    2, 3,7, 8-TCDD," Environ. Sci. Res. 12,  1978, 381-388.
11.  IARC.,  "Chlorinated dibenzodioxins In: Monographs on the Evalua-
    tion of the Carcinogenic Risk  of Chemicals to Man," 15, 1977,41-102.
12.  Jirasek, L.,  Kalensky, J., Kubec, K., Pazderova, J. and Lukas, E.,
    "Chlorakne, porphyria  cutanea tarda  und  andere  Intoxikationen
    durchHerbizide,"//awtarz/.27, 1976, 328-333.
13.  Kimmig, J. and Schulz, K.H., "Occupational acne (chloracne) caused
    by chlorinated aromatic cyclic ethers," Dermatologica  115,  1957,
    540-546.
14.  Kociba, R.J., Keyes, D.G., Beyer, J.E. et al.,  "Results of a two year
    chronic  toxicity and oncogenici'ty study of 2, 3, 7, 8-TCDD in rats,"
    Toxicol. Appl. Pharmacol. 46, 1978, 279-303.
15.  May, C., "Chloracne from the accidental production of tetrachloro-
    dibenzodioxin," Brit. J. Ind. Med; 30, 1973,276-283.
16.  Poland, A.P., Smith, D., Metter,  G. and Fossick,  P., "A health
    survey of workers in a 2, 4-D and 2, 4, 5-T plant," Arch.  Environ.
    Health 22, 1971,316-327.
17.  Reggiani, G., "Acute human exposure  to TCDD in Seveso, Italy,"
    J. Toxicol. and Environ. Health 6, 1980,27-43.
18.  Strik, J.J.T.W.A. and Koeman, J.H., Chemical Porphyria in Man.
    Elsevier/North-Holland Biomedical Press, Amsterdam, 1979.
19.  Thigpen, J.E., Faith, R.E., McConnell, E.E.  and Moore, J.A., "In-
    creased  susceptibility to bacterial infection as a sequela of exposure
    to 2, 3, 7, 8-TCDD," Infec. Immunol. 12, 1975, 1319-1324.
20.  Van Miller, J.P., Lalich, J.J. and Allen,  J.R.,  "Increased incidence of
    neoplasms in rats exposed to low levels of 2, 3, 7, 8-TCDD," Chemo-
    spherelO, 1977,625-632.
      RISK ASSESSMENT/DECISION ANALYSIS
                                                                                                                                289

-------
THE  APPLICATION OF  QUANTITATIVE  RISK  ASSESSMENT
              TO ASSIST  IN SELECTING  COST-EFFECTIVE
                              REMEDIAL ALTERNATIVES

                                       LAWRENCE J. PARTRIDGE
                                        Camp Dresser & McKee Inc.
                                           Boston, Massachusetts
INTRODUCTION

  The Western Sand and Gravel (WSG) site in Burrillville, Rhode
Island was ranked 128 on the National Priorities List published in
September 1983. Hazardous wastes are migrating from the site via
groundwater and  surface  water and  are contaminating nearby
domestic drinking water wells, a major groundwater aquifer and
the Slatersville Reservoir which is designated as a Class B surface
water body. The objective of this study was to evaluate the feasi-
bility for groundwater remediation and to identify cost-effective re-
medial  alternatives which would minimize public health and en-
vironmental risks.

SITE HISTORY

  The WSG site (Fig. 1) is located in a semi-rural area of Burrill-
ville, Rhode Island, adjacent to the Douglas Pike and close to the
North Smithfield town line. This site was originally a gravel min-
ing operation, however, beginning in 1975, a portion of the 12 acre
site was used for the disposal of septage and chemical wastes. The
wastes were dumped into unlined seepage lagoons and allowed to
infiltrate into the soil, percolate through the permeable soils and
enter the groundwater.
  The available hazardous waste manifest records indicate that
approximately 470,000 gal  of chemical wastes were dumped  at
WSG from May 1978 through April 1979. No records were avail-
able to  quantify the wastes dumped at the site prior to May 1978.
The site was closed in May 1979.
  The USEPA undertook an emergency cleanup program at WSG
in 1980. Liquid chemical wastes and sludges were removed from the
site and sent to licensed disposal facilities.  It was estimated that
60,000 gal of waste were removed from WSG during the cleanup
program while approximately 400,000 gal of waste remained.

HYDROGEOLOGIC SETTING

  The topography of WSG generally results from a combination of
Ice Age land forms and modern day gravel mining operations. Sub-
surface  natural deposits are stratified drift having a glacio-fluvial
and ice  contact origin. These deposits consist of gravelly or sandy
sediments which were deposited by meltwater streams associated
with the deglaciation of the region. These sediments are a part of
the Slatersville Aquifer.' Strata on either side of Tarkiln Brook
are  finer  grained,  slightly  younger  glacial  sediments primarily
represented by stratified fine sands.
  The WSG site is located on the east side of an irregular bedrock
valley with depth to bedrock varying from 30 to 70 ft below the
land surface. The apparent centerline of the bedrock valley tends
north-south in a direction paralleling the Tarkiln Brook. The bed-
                                                  rock trough is not evident near the Slatersville Reservoir, and there
                                                  is a dip in the bedrock surface to the southwest.
                                                    The source of groundwater on-site is infiltration of precipita-
                                                  tion and  lateral groundwater flow  from the south  and east.
                                                  Groundwater flow from the site is to the  west and north toward
                                                  Tarkiln Brook  and  the  Slatersville Reservoir. During its flow
                                                  through the site toward the Reservoir, the Tarkiln Brook is re-
                                                  charged at the rate of 12,000 gal/day.

                                                  EXISTING CONTAMINATION

                                                    Large quantities of hazardous wastes still remain at WSG. These
                                                  wastes are found in the groundwater, in the surface water of the
                                                  Tarkiln Brook and in surface and subsurface soils.  There is also a
                                                  layer  of nonaqueous phase liquids floating on the groundwater
                                                  surface in the vicinity of the groundwater  pumping system at  the
                                                  site. These hazardous  wastes  (Table  1) represent a  continuing
                                                  source of contamination for both groundwater and surface water.
                                                    There is, however, no indication that  airborne emissions of
                                                  vapor or paniculate from WSG are affecting the areas surround-
                                                  ing the site.  No  volatile emissions  were detected  at  the  site
                                                  perimeter  during a site survey using direct reading  instruments
                                                  (HNu Photoionizaiion detector).

                                                  STATUS OF REMEDIAL ACTIVITIES

                                                    The Rhode Island Department of Environmental Management
                                                  (R1DEM) has sponsored  several studies to evaluate on-site con-
                                                  ditions at  WSG and implemented programs to reduce the migra-
                                                  tion of hazardous  wastes from the site. Work programs to date
                                                  have included:
                                                  •Hydrogeologic investigations
                                                  •Bedrock contamination studies
                                                  •Treatment feasibility studies
                                                  •Design and installation of an emergency groundwater recircula-
                                                   tion system
                                                  •Performance evaluation of the emergency groundwater recircu-
                                                   lation system.
                                                    The existing  emergency groundwater pumping  system, which
                                                  consists of one extraction well and five recharge  chambers,  was
                                                  installed in the fall of 1982. The purpose of the system was to pro-
                                                  vide emergency containment  of the contaminated groundwater
                                                  plume and minimize its migration off-site. A skimmer pump con-
                                                  nected to the system has also captured approximately 1,000 gal of
                                                  the non-aqueous phase liquids which are floating on the ground-
                                                  water surface. These liquids are being stored on-site in a holding
                                                  tank prior to shipment off-site for treatment and disposal.
290
RISK ASSESSMENT/DECISION ANALYSIS

-------
                                                           Figure 1
                                                   Western Sand and Gravel Site
  This current pumping system has been operational for approx-
imately 15 months and has pumped nearly 9,000,000 gal of ground-
water for redistribution to the recharge chambers. There have been
some periods of pumping interruption and procedures are being
implemented to improve the performance of the system. The sys-
tem has effectively captured the non-aqueous phase liquids float-
ing on the groundwater surface.  However, a review of the avail-
able downgradiant monitoring data before and after installation of
the pumping system does not show appreciable change in contam-
inant levels measured over time.

CONTAMINANT MIGRATION

  Samples  collected during this  program  plus  additional data
collected previously reveal  that  both groundwater and  surface
water at WSG are contaminated.  These contaminated waters flow
in a northerly direction from WSG and  discharge to the  Slaters-
ville  Reservoir (Fig. 2). While the Slatersville Reservoir is not a
source of drinking water and  is  used only for recreational pur-
poses, the Reservoir does overlie the Slatersville Aquifer which has
been identified by the U.S.  Geological Survey (USGS) as a major
groundwater aquifer in Rhode Island. The USGS indicated that
induced infiltration from the Slatersville Reservoir would be the
principal source of recharge  to a well field installed in this aquifer.
  An analysis of the horizontal and vertical gradients near the
Slatersville Reservoir  indicates  that approximately  80%  of the
contaminated groundwater discharges to the Reservoir while 20%
recharges the Aquifer. Total contaminant loading to the Reservoir
is approximately 3 gal/day of volatile organics assuming a ground-
water flow of 9,000 gal/day with a worst case volatile organic con-
tent of 410 ppm based upon observations at well GZ-3 (Table 2).
The dilution capacity of the Reservoir in the area of the contam-
inated discharge was estimated  at 30 million gallons yielding an
effective concentration of 0.1  mg/1 of total volatile organics. The
concentration of specific chemicals  would be appreciably lower.
Approximately 0.7 gal/day  of volatile organics would flow to the
Aquifer which, in turn, discharges to the Reservoir. Given the cur-
rent and projected recreational  use  of the Reservoir, there is no
acute or chronic public health risk associated with exposure to the
toxic  substances via ingestion of or  dermal contact with contam-
inated surface water in the Reservoir.
  Several residential drinking water wells adjacent to WSG are
contaminated with low levels  of hazardous chemicals. These wells
(Fig.  2) are being monitored by the Rhode Island Department of
Health (DOH) as part of their program for evaluating groundwater
quality in the area adjacent to WSG.  There are currently seven
domestic drinking  water  wells which have shown the presence of
                                                                         RISK ASSESSMENT/DECISION ANALYSIS
                                                          291

-------
 1,2-Dlchlorobenzene
 Napthalene
 Nltrodiphenyl anlne
 Pencachlorophenol
 Anthracene
 Dl-Bucylphchalace
 Bis-2-ethyl hexyl phthalate

 Other Non Priority Pollutants
                                                           Table 1
                                             Residual On-Slle Contamination at WSG

                                     Concentration In Selected Media (ppb)
Chemical
Hethylene chloride'
1 , 1-Dichloroethane
Trans-1 ,2-Dlchloroethylene*
Chlorof or«
I.I, l-Trlchloroethane
Trichloroethylene*
Tetrachlorocthylene
Chlorobenzene
Benzene*
Toluene
Ethylbenzene
Xylenes
PCB's
Contaminated
SolU
Aug. 1982
21.000
3.000
1 .000

13,000
35.000
60.000
156.000
2,000
20,000
69,000
355.000

Non-Aqueous1
Phase Liquid*
Dec. 1982
140,000
20,000
15.000
120,000
4,900.000
4,100,000
3.800.000
25.500.000
400.000
1 S3. 000, 000
20,300.000
78,400,000
73,000
Contaminated Groundwater
Nov. 1983
OW-1
28.400
1.680
7,660
4,000
19.100
10.300
1.560
17.000
510
37,800
5.460
19.600

E2-3
46.600
540
4.080
3,140
30,000
13.500
2.400
23.300
830
231.000
7.780
39.000

                             August 1983

                                4.900
                                2,600
                               47,000
                                2,300
                                2,000
                                6.500
                               60.000

                              316.000
360.000
 63.000
 1. Based upon GC and OC/MS uulym.
                                                                                                              •Suspect CaroDojoM
                                                           Table 2
                                      Analysis of Groundwater Samples, Western Sand and Gravel

                                    Concentrations of Various Sampling Sites (pg/1)
Parameter
Volatile Organic Compounds
•ethylene chloride
1 , 1-dlchloroethy lent
1 . 1 -dlch loroe thane
lraus-l,2-j|ch loroethyl ene
ch loruf M>
190
720
10 pg/l

SS-3

u
NO
6
98
2
1
SO
ND
2
ND
ND
20
ND
ND
ND
ND
1
ND
28
ND
1M
22
10)

ND - nc
SS-5

ND
ND
21
2)
2
2
J
ND
1
ND
ND
ND
NO
ND
ND
NO
NO
ND
1
ND
<.S
29
11)

>nc detected
292
RISK ASSESSMENT/DECISION ANALYSIS

-------
Parameter
                                          GZ4-1
   Table 2 Continued

GZ4-2            GZ4-3
                                                                                                GZ5-1
GZ5-2
                                                                        GZ5-3
Volatile Organic Compounds
•ethylene chloride
1 ,1-dlchloroethylene
1 ,1-dlchloroethane
trans-l,2-dlchloroethylene
chlorof or«
1,2-dlchloroe thane
] , 1 ,1-trlchloroethane
carbon tetrachlorlde
brDaodlchloroBe thane
1,2-dlchloropropene
t rans-1 , 2-dlchloropropene
trlchloroethylene
1,1,2-trlchloroethane
cla-1 , 3-dlchloropropene
dlbroBochloro*e thane
bronofom
te t rach 1 oroethy I ene
1,1,2, 2- tet rach 1 oroe thane
chlorobenzene
benzene
toluene
ethylbenzene
xylene
Detection Limit
NO
NU
NU
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
NU
ND
ND
NU
ND
ND
NU
NU
1 pg/1
ND
ND
3
55
ND
ND
ND
ND
ND
ND
ND
1
ND
ND
ND
ND
ND
ND
NU
ND
t,
ND
ND
1 Pg/1
ND
16
191,
1,10
t.
1,
'lit
3
ND
ND
ND
10
ND
ND
2
ND
5
ND
13
35
<>3
Vt
170
lMg/1
ND
ND
ND
ND
NU
ND
ND
ND
ND
ND
NU
ND
ND
ND
ND
ND
ND
ND
ND
ND
5
ND
ND
1 Mg/1
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
1 Mg/1
3
2
K
28
ND
ND
4,
ND
ND
ND
ND
ND
ND
ND
ND
1
ND
ND
ND
3
5
It
'21,
1 Pg/l
 1. EPA Ambient Water Quality Criteria for Human Health.
                                                                                                           •Suspect Human Carcinogens 10  * Cancer Risk.
                              \
                    X   /
                          bOoO-BOOO Ft /(Uy


                                 Contaminated Domestic Wells


                                 Monitoring Well Locations


                                 Surface Sampling Locations


                            Scale:! inch = 400 ft.
                                                                   Figure 2
                                                         Hydrogeologic Setting at WSG
                                                                                 RISK ASSESSMENT/DECISION ANALYSIS
                                                                             293

-------
                ufnn imrr •*_*
                tuu Kin 1  }
                                        I O  Ul.l

                                         . /u
                                                          Figure 3
                                                  Physical Conditions al WSG
organic chemical contamination at least once in their sampling his-
tory. Four of these wells have shown contamination each time they
were sampled over the past 30 months and three are located within
the Slatersville Aquifer (Fig. 2).
  A volatile organic priority pollutant analysis of groundwater
samples collected at WSG and adjacent areas is found in Table 2.
Table 3 shows the USEPA Quality Criteria for Human Health and
Aquatic Toxicity and the designation of those contaminants which
are suspect human carcinogens. Several of the wells listed in Table
2 are  multi-level, and these provide a profile on water quality at
different depths  below the ground surface.  For example, samples
designated GZ1-1, GZ1-2 and GZ1-3 are taken from the same well
location. GZ1-1  represents a  sample collected 7.5 ft below the
ground surface,  GZ1-2 indicates  a sample  taken 45 ft below the
ground surface  and GZ1-3 was  75  ft below  the surface. Similar
multi-level  sampling  configurations  exist in  wells  E3, GZ4 and
GZ5.  A review of the data indicates that contaminant concentra-
tions are highest at the deepest location within a given well. For ex-
ample, locations E3-2,  GZ1-3, GZ4-3 and  GZ5-3, which are the
deepest sampling points in each well, all show levels of contamina-
tion higher than at shallower sampling locations in the same wells.
Wells GZ4 and GZ5 show  upward groundwater gradients at  their
deepest sampling locations. No gradient data are available for wells
E3andGZl.
  The individual groundwater sampling locations shown on Figure
2 indicate the northerly movement of contamination from WSG to-
ward  the Slatersville Reservoir.  Residential sampling well RW-1
in Table 2 has the highest level of contamination among the domes-
tic wells. This well is no longer used for drinking water, and the
resident has chosen  to use bottled water. Also, surface sampling
locations SS-3 and SS-5 represent samples collected in the Tarkiln
Brook. Total volatile organics  measured at SS-3 and SS-5  were
approximately 530 and 260 yg/1, respectively. The concentration
for some chemicals in the stream  samples exceeded USEPA water
quality criteria for drinking water but did not pose an acute or
chronic toxicity hazard for aquatic species.
                                                       INITIAL ASSESSMENT OF PUBLIC HEALTH RISK
                                                         The estimates for public health risk are based upon the results
                                                       from the monitoring of groundwater and surface water contamina-
                                                       tion at WSG. Additional considerations relate to the quantities of
                                                       hazardous  wastes which remain at the site and the possibility for
                                                       their release to groundwater and surface water.
                                                         The public health risk  assessment is  based  upon the contam-
                                                       inant concentration levels measured in well GZ1 which is the mon-
                                                       itoring well closest to impacted residential wells. This estimate of
                                                       risk presented in Table 4 is based upon USEPA Water Quality
                                                       Criteria for Human Health. The total risk estimate, assuming that
                                                       carcinogenic risks are additive, is equivalent to 310 x 10~* based
                                                       upon a lifetime ingestion of drinking water containing  the chem-
                                                       icals at concentrations shown in Table 4.
                                                       ASSESSMENT OF REMEDIAL ALTERNATIVES

                                                         The data presented in Table 4 indicate that  the chronic inges-
                                                       tion of contaminated groundwater poses a lifetime cancer risk of
                                                       310 x 10~' based upon USEPA water quality criteria. This cancer
                                                       risk could be higher if  1,1  dichloroethane and trans-1,2-dichloro-
                                                       ethylene are shown  upon  further study to  be  carcinogenic. The
                                                       analysis proceeds to evaluate the relative effectiveness of remedial
                                                       alternatives with respect to reducing contaminant migration from
                                                       the site.
                                                         A two-dimensional cross-section flow and solute transport model
                                                       was used  to examine  the effectiveness of the various remedial
                                                       alternatives. The physical representation of the site,  including the
                                                       geometry and physical (hydrogeological) conditions, is shown on
                                                       Figure 3. This grid system for the specific cross section was  selected
                                                       to generally coincide with the groundwater flow line which parallels
                                                       flow in the Tarkiln Brook. The water table elevation  for the upper
                                                       laver was specified initially and held constant during  the period of
                                                       simulation. The bottom and lateral edges of the grid were defined
                                                       to be no-flow boundaries.  For the left edge  of the grid, this repre-
294
RISK ASSESSMENT/DECISION ANALYSIS

-------
                                                               Table3
                                                Selected Water Quality Criteria Information
Compound
nujtliylene chloride
1,1 d ichloroethylene
1, 1-diehloroettiane
lrjn*-l , 2-diirliloroetliylene
USEPA Ambient
Water Quality Criteria
for Human Health,,
(pg/L)
O.!9dk
O.OJd
insuff Icient data
insuf f ic iu-nt data
Calculated Drinking
Water Criteria!)
1
jig/1) would be expected to reach grid  block 20 (Fig. 3) within 5
years and equilibrium concentrations of approximately 2,700 ;ig/l
(or 27% of the  source strength) could be expected within  15 years
for no-action alternative. The  calculated contamination distribu-
tions at the end of 10 and 15 years are shown in Figure 4.
  The  contaminant plume for the retarded species (Fig. 5) is much
less extensive at the end of 15 years. However, it still effects a sig-
                                             g. Based on organoleptic data, based on health effects data the criteria would be 488 pg/L
                                             h. For suspected carcinogens at 10 -' risk
                                             i. Proposed recommended maximum contaminant levels
                                             j. MCL for total trihalomethanes
                                             k. For total halomethanes

                                             nificant area under the no-action alternative. It is important to
                                             recognize that  these models  are only approximate and  that  the
                                             actual extent of plume migration at WSG is probably represented
                                             by a  situation  intermediate between the predictions  of the two
                                             models.
                                               A second simulation was made  to estimate the natural restor-
                                             ation capabilities of the groundwater flow system. An initial con-
                                             taminant distribution throughout the aquifer system was specified.
                                             The specified distribution was similar to the 10-year calculated dis-
                                             tribution for the mobile chemical  species of the previous simula-
                                             tions. The source term concentration was defined to be zero, which
                                             represents  either complete encapsulation or  removal  of  the con-
                                             taminant source.
                                               The purpose  of  the  simulation  was  to estimate the minimum
                                             time the aquifer would remain contaminated after on-site remedial
                                             action.  The model  results indicate that five  year  contamination
                                             levels immediately  below the site would be below detection (< 1
                                             jig/1) and the highest concentrations (approximately 3,000 jig/1)
                                             would be in the vicinity of GZ-5. After 10 years, the highest con-
                                             centration levels would be approximately 200 jug/1 in the vicinity of
                                             GZ-1. It would take approximately 15 years for the contamina-
                                             tion levels  in the vicinity of GZ-1  to drop below 1,000 jig/1. Fig-
                                             ure 6 illustrates the change in position of the l.OOOjug/1 iso-contour
                                             at five-year intervals. The results  of these  analyses  indicate that
                                             with natural cleansing,  groundwater in the vicinity of  GZ-1  re-
                                             mains contaminated (at levels greater than 1,000 jig/1) for at least
                                             15 years and that between 15  and  30 years are required  for levels
                                             to decrease significantly below 1,000 jig/1.
                                               The results from the preliminary model runs were used to estab-
                                             lish the initial conditions  for  the  evaluation  of remedial alterna-
                                             tives. Initial conditions for the hydrogeologic system relate to the
                                                                             RISK ASSESSMENT/DECISION ANALYSIS       295

-------
                                                                                        czi
A
1
5
9
use
Source
Strength
10,000 ppb

w
~T-V
"r>
'ii\ *
1
•H
>n
\1 * I
-1

lorlzo
Unrt if

\|
v°,
0.

3
n
nt
n l
T
.9
4l
\ *i
V
> s
O.A
J*^-
It
al
0 S
2
\'
S
i
1
i

\
\\
5
Sea
r-al


L\
\
k
le
P

"^

k
^

1 1
1 1
~\





^







0.4



— ~_


N-
0.4


_^-=
	



\
^x
10
nch = 200 ft.
iit-li = ?2() fi




_...


\
^._



O.I 	




	
/






•

	








N.




















\
\
I











—








\
\
\
\



^

15 20




\
N
t
/
f


	


















25
10 years
15 years
                                                          Figure 4
       Fractional Concentration of Mobile Species after 10 and 15 Years of Source Loading at 10,000 ppb. X and Y Axes Correspond to Figure 3.
    wsc
    Source
    Slrengtl
    100.000
        pph
                                                                                        CZI
 1
                               10
                                                                                      20
25
llorlzonl.il  Scnle   1  Inch   200 ft.
Vc-rl lr.il  Sc iile     1  Ini'h -   20 f I  .
                                                           Figure 5
            Fractional Concentration of Retarded Species after 15 Years of Loading at 100,000 ppb. X and Y Axes Correspond to Figure 3.
296      RISK ASSESSMENT/DECISION ANALYSIS

-------
                                                               Table 4
                                                      Baseline Level of Health Risk
Well Concentrations (ug/1)
1982
GZ1-2 GZ1-3
1 , 1-dichlorootliane
1 , J-dicliloroethylene
trans-1 ,2-dichloroethyleno
chloroform
1 ,2-dichloroetliane
1,1, l-lrichloroetliane
chlorobenzene
benzene
toluene
ethylbenzene
xylenes
TOTAL
37
ND
110
NL>
3
14
2
19
25
31
153
394
131
2
128
ND
14
31
12
52
236
252
1000
1856
1983
G/l-2
184
10
398
ND
2
12
23
10
87
32
120
868
GZ1-3
226
14
202
51

66
64
38
440
190
720
1946
Average USEPA Water Quality
Concentration Criteria for Health
(ug/L) 
-------
                            Figure 7
           Range of Total System Costs Versus Level of Risk
reduction  in source strength that  is anticipated following the im-
plementation of the various remedial alternative considered for the
WSG site.
  The  initial conditions regarding flow rate at the site and con-
taminant loading rate for the remedial alternatives considered in
this study are shown in Table 5. Simulation runs were conducted
for each  remedial  alternative,  and final steady-state conditions
were measured to evaluate the effectiveness of the remedial alter-
native. Final steady-state contaminant levels under each remedial
alternative were then compared with the results under a no-action
alternative. The reduction in the level of contamination was then
accepted as a measure of the system efficiency.
  The  removal efficiency was then employed to evaluate the level
of risk reduction for each  remedial alternative. Risk  reduction is
based upon the removal  of suspect human carcinogens from the
groundwater and a concomitant reduction in the level of chronic
ingestion. The level of risk associated with each remedial alterna-
tive is presented  in Table 5. Analysis shows that the greatest level
of risk reduction is achieved through the implementation of the
slurry  wall  with  treatment  and  recirculation  of the  pumped

Do Nothing

Slurry Wai 1

Slurry Wai 1 6
Treatment

Hurry Wai 1 i
Internal Rec i rcu-
lation t Treatmen

Hydrodnamlc Isola

Hy^rpdynAml ^ l*o»
latlon I (Treat men

:xpnddd Hyldrodypm

3uy-0ut



to

10

10


)



J

)


}

10



:o

10

)


j



!

3


3

1



10

10

J


!



3

J


J

10



1

)

10


10



10

10


10

]



1

10

10


10



13

IU


10

10



1

10

3


)



10

J


)

J



10

10

s


)



10

)


3

I



10

10

5


5



:o

1


l

10



1

3

10


10



3

to


3

'.



1

10

10


10



10

10


10

1



1

10

10


10



10

10


10

10



1

10

10


10



10

10


:o

10



i

10

10


to



;o

10


10

3



1

10

10


10



LO

10


10

3



10

10

10


10



10

:o


10

3


                                                                                                 I JO
                                                                                                 120
                                                                                                  IV
                                                            Figure 8
                                              Second Level Analysis of Remedial Options
298       RISK ASSESSMENT/DECISION ANALYSIS

-------
                            Tables
            Risk Levels Associated with Remedial Options
               Flow Rate    Contaminant    Removal
               at Site      Loading       Efficiency    Risk
Remedial Option   (gal/day)    (pg/1)         (%)        Level x 10 ~«
Do-Nothing
Slurry Wall
Slurry Wall&
& Treatment
Slurry Wall &
Internal Recircula-
tion & Treatment
Hydrodynamic
Isolation
Hydrodynamic
Isolation &
Treatment
Expanded
Hydrodynamic
9,000
900

900


300

1,300


1,300

900
10,000
1,000

30


15

1,440


144

1,000
0
90

99


99 +

86


98

90
310
31

3


3

43


6

31
effluent. The slurry wall with treatment and no recirculation and
the groundwater pumping system also demonstrated relatively high
removal efficiencies.

 IDENTIFICATION OF COST EFFECTIVE REMEDIAL
 ALTERNATIVES
   The levels of risk reduction information in Table 5 was com-
 bined with net present value for total system costs to assess the rela-
 tive effectiveness of the remedial alternatives. This information for
 the eight options considered in this program is shown in Figure 7.
   The information shown in Figure 7 provides a basis for estab-
 lishing a feasible domain regarding the level of public health risk
 and costs for the remedial alternatives.
  A variety of scientific literature suggests that humans are gen-
erally willing to accept risks to health in the range of 10~5 to 10~'.
Given this criteria, it appears that remedial Alternatives 1,2, 5, 7
and  8 should be excluded from further considerations. A  second
major consideration regarding the remedial alternatives relates to
the net present value of capital and operating costs for the life of
the system. The total system costs as presented in Figure 7  consist
of a range which reflects a measure of uncertainty. For example,
the total system costs for Alternative 3, the slurry wall and treat-
ment system ranged from $7,800,000-$8,900,000. This is typical of
the cost variations which can be anticipated in this type of analysis.
Based upon the cost analysis, it appears that systems 3 and 6 repre-
sent the most cost-effective remedial alternatives with respect to
risk reduction.
  Public health risk and system costs are two parameters which
can be employed to evaluate remedial alternatives. However, there
are a variety of other screening criteria which should be employed
to select viable remedial systems. These criteria, shown in Figure 8,
were ranked from 1-10 for each of the remedial systems considered
in this analysis. Each of the  secondary screening criteria is un-
weighted and the total score for each system is shown in the right
column. The results from the secondary screening indicate that re-
medial systems 3 and 6 score near the top with respect to  overall
ranking of the remedial alternatives.
  In conclusion, the application of a quantitative risk assessment in
conjunction with considerations for cost and secondary screening
criteria provides a basis for evaluating remedial alternatives  at haz-
ardous waste disposal sites. This methodology provides a basis for
explicitly considering public health risk in the evaluation of remed-
ial alternatives.

REFERENCES
1.  Johnston, H.E. and Dickerman, D.C.,  Availability of Ground-water in
   the Branch River Basin, Providence  County,  Rhode  Island, U.S.
   Geological Survey, Water Resources Investigations, 1974, 18-24.
                                                                         RISK ASSESSMENT/DECISION ANALYSIS
                                                         299

-------
         APPROACHES TO COMPUTER RISK  ANALYSIS AT
              UNCONTROLLED HAZARDOUS  WASTE  SITES

                                                 BAXTER JONES
                                                  KEN KOLSKY
                                                 ICF Incorporated
                                                 Washington, D.C.
INTRODUCTION
  In this paper, the authors outline an approach to the assessment
of health risks at uncontrolled hazardous waste sites, and they
discuss several  applications of computers in the risk analysis pro-
cess. An integrated risk analysis model currently being developed
for hazardous waste facilities is then described to illustrate some of
the principles of computerized risk assessment. Finally, some of the
advantages and disadvantages of using computers in risk anslysis
are summarized. Before proceeding,  however, three basic questions
should be addressed.
What is Risk Analysis?

  Risk analysis is defined here, in the context of hazardous waste
sites, as "the systematic scientific characterization of the probabili-
ties and types  of adverse  effects that may result  from  chemical
releases at the site." The authors focus on risks of chronic human
health effects, although risk analysis can also be used to evaluate
ecological and  environmental effects, health effects  of acute  ex-
posures or other types of adverse  consequences.  To the degree
possible, risk  analysis  is  independent of political, legal  and
economic considerations  that  must be factored  into  decision-
making. This distinction between the scientific evaluation of risk
and the management of risk is consistent with the recent recom-
mendation of the National Academy of Sciences (NAS)'  and with
current USER A policy.2 The approaches and models described in
this paper are  risk analysis tools which must be integrated with
other sources of information in site  decision-making.

Why Use Risk  Analysis?

  There are a number of reasons to use risk analysis at hazardous
waste sites.  In  fact, rational decisions about remedial actions at
most sites cannot be made in the absence of health risk  analysis,
although the form and extent of the analysis may vary widely, even
to the point of being unrecognizable as risk analysis.  The authors
would argue that even the use of existing environmental criteria or
standards to guide actions at a site is a form of  combined risk
analysis/risk  management. Thus, the question is not so much
whether to use risk analysis in remedial site management,  but what
form the analysis should take and how comprehensive it should be.
  Specific applications of  risk analysis at individual uncontrolled
hazardous waste sites include the following:
•Evaluation and comparison of site remedial action alternatives, on
 a health basis
•Final design specification  for a selected remedial alternative
•Characterization of baseline site risks (i.e., risks in the absence of
 remedial action)
                                                     •Characterization of residual site risks (i.e., risks existing follow-
                                                      ing the remedial action)
                                                     •Evaluation of health benefits, in terms of risk reduction, result-
                                                      ing from a remedial action
                                                       Thus, risk analysis can play a role in demonstrating the need for
                                                     action at a site, in selecting and designing a remedial action for the
                                                     site and in evaluating the effectiveness of the action. Risk analysis
                                                     can also be applied to questions concerning multiple sites, such as
                                                     site ranking, establishment of cleanup priorities and evaluation of
                                                     overall remedial program benefits.
                                                     What are the Drawbacks to Risk Analysis?

                                                       If systematic risk analysis is so useful at uncontrolled hazardous
                                                     waste sites, then why has it not been more widely used? The authors
                                                     believe that there are three primary reasons for its lack of utiliza-
                                                     tion:
                                                     •The unavoidable  analytical uncertainties  involved in environ-
                                                      mental risk analysis
                                                     •The substantial site-specific data requirements for a defensible
                                                      analysis
                                                     •The perception that risk analysis will slow the remedial process
                                                       The authors do not dismiss these concerns, and in fact do point
                                                     them out as a cautionary note. Nevertheless, the authors believe
                                                     these concerns to be  counterbalanced by the potential benefits of
                                                     risk analysis at a site.
                                                       Uncertainty is perhaps the foremost certainty in environmental
                                                     risk analysis, and site decision-makers seeking precise risk estimates
                                                     will be disappointed. However, by estimating the reasonable ranges
                                                     of risk estimates and by using techniques such as sensitivity an
                                                     analysis and worst-case assessment, one can produce useful risk
                                                     analyses despite the uncertainties. The data needs for risk analysis
                                                     are considerable, but these  can be met during the site characteriza-
                                                     tion process at most sites. For this to happen, the risk analysts must
                                                     be involved in planning and monitoring site characterization ac-
                                                     tivities. As  for the perception of slowing site remedial progress, the
                                                     authors do not  think this will be significant  if risk analysis is in-
                                                     tegrated  into the  overall  remedial  process. It can almost be
                                                     guaranteed, however, that  a risk  analysis tacked on late in the
                                                     remedial process will prove unsatisfactory and will slow site pro-
                                                     grejs.
                                                       In the remainder of this paper, the authors present a general risk
                                                     analysis approach for hazardous waste sites,  describe various uses
                                                     for computers in the process, present an example of an integrated
                                                     risk analysis model and summarize the advantages and disadvan-
                                                     tages of risk modeling using computers.'
300
RISK ASSESSMENT/DECISION ANALYSIS

-------
                Cha racter i ze Chenica I s
                      at S i te
            1.   Identify chemicals present
            2.   Detemi ne quant i ty, fora,
                location or chemicals
            3.  Select  in
                a na I y s i s






                                                        Identify types of toxicity
                                                     2.  Determine dose-response/
                                                        thresholds
ASS
jss Exposure
                                                     1.  Identify exposure pathways.
                                                     2.  Characterize populations-
                                                        at-risk
                                                     3.  Estimate  release probabili-
                                                        ties/quant i t tes
                                                        Evaluate environmental
                                                        fate/transport
                                                     5.  Model human intake/adsorption
                                                                                                  1.  Conbine exposure and
                                                                                                     toxicity information
                                                                                                  2.  Develop various  risk
                                                                                                     measures
                                                             Figure 1
                                           Risk Analysis Approach for Hazardous Waste Sites
GENERAL RISK ANALYSIS APPROACH
  The approach described for risk analysis at uncontrolled hazar-
dous  waste  sites  is  adapted from  traditional  risk  assessment
methods used for pesticides, occupational agents, food additives
and  drugs.  However,  it  takes  into  account some  of  the
characteristics of waste sites that make risk analysis more difficult:
multiple  chemicals;  multiple  release  sources  and  exposure
pathways; multiple exposed  populations; and separation of release
sources and receptors in time and space. The approach is consistent
with the chemical risk assessment principles outlined recently by the
NAS.1 The approach is only briefly described here, as the focus of
this paper is computer applications.
  The risk analysis approach consists of four steps that roughly
parallel the steps outlined by NAS. The first step is characterization
of the toxic chemicals present at the waste site. The next two steps
are chemical toxicity assessment and exposure assessment. Finally,
the exposure and toxicity information is integrated to characterize
the risks. Figure 1 is a flow  chart showing the overall risk analysis
approach.

Characterize Chemicals
  The first step in risk analysis is to characterize the toxic chemicals
present at  the waste site. In addition to simply identifying  the
chemicals, their locations at the site and the quantities present must
be estimated. This step is  based on chemical analysis data for the
site and in some cases can be supplemented by site disposal records.
Because many waste  sites  may contain more  than  100 toxic
chemicals,  indicator chemicals  should be  selected  for the  risk
analysis. It is impractical  and  unnecessary  to  include  a large
number of chemicals in the risk analysis; a few indicators of the  risk
are selected based on their toxicity, environmental mobility, per-
sistence and the quantity present.

Assess Toxicity
  In this step, the inherent  toxicity of each indicator  chemical is
assessed. The types of toxic effects, the levels at which these effects
occur and the dose-response relationships must be determined for
each chemical. The precise form of the quantitative data required
will depend on the  dose-response model being used. Because the
assessment of inherent toxicity does not require site-specific data,
this step can be done prior to the actual risk analysis.
  Usually, this assessment is  based on human  epidemiological
studies or experimental animal studies.  In practice, many of the
chemicals encountered will have toxicity  assessments available and
will not need to be reassessed. For example, the USEPA's Car-
cinogen Assessment Group has  derived dose-response data for
many suspected carcinogens,3 and the USEPA and the FDA have
determined acceptable daily intake  (ADI) values for many non-
carcinogens.4 If further assessment is required,  however, the ser-
vices of an experienced toxicologist will be needed to complete this
step.

                            Table 1
               Hypothetical Exposure Pathway Matrix

Release Source
Site leachate
Site leachate
Abatement plant
effluent
WTP Volatilization
Wellhead treatment
volatilization
Abatement plant
volatilization
Exposure
Pathway
Release/
Transport
Medium
Groundwater
Groundwater
Surface water
Air
Air
Air

Point of
Exposure
Public drinking water
Private drinking water
Downstream
swimming
Nearest residence
Nearest residence
Nearest residence
Remedial
Alternative
1 2 3
X X
XXX
X
X X
X
X

4
X
X
X
X

X
Assess Exposures
  This is typically the most complex and time-consuming part of
the analysis. Exposures of people affected by the waste site must be
quantified for all indicator chemicals. Although environmental and
human monitoring data can be extremely useful  if available, this
step almost always requires modeling of chemical releases from the
site and transport  to the exposure point. Some type of modeling,
which can vary from a series of logical assumptions to a full multi-
media simulation of the waste site, is necessary whenever exposures
must be estimated into the future.
  The exposure assessment can be divided into several activities.
First, an exposure pathways analysis must be done to identify: (1)
the significant sources of potential chemical release to the environ-
                                                                          RISK ASSESSMENT/DECISION ANALYSIS
                                                          301

-------
ment, (2) the release and transport media (e.g., air, groundwater)
for each source, (3) points of potential human exposure and (4) the
exposure medium and exposure route. A hypothetical exposure
pathway  summary  for an  uncontrolled hazardous  waste site  is
shown in Table 1. Each complete exposure pathway that is judged
significant  represents a chain of events for which  chemical ex-
posures must be quantified. The population-at-risk for each ex-
posure point (or area) also should be quantified.
  For each significant exposure pathway, chemical release from the
source must be projected, preferably in the form of release  profiles
for individual  chemicals.  A  release probability must  also  be
estimated for stochastic sources of release. Typically, the  estima-
tion of chemical releases is one of the more problematic steps in risk
analysis for hazardous waste sites. Following the projection of
release profiles for  a specific source, the fate and  transport of
chemicals  between the source and any exposure points must  be
evaluated. This analysis can be done using simple dilution, disper-
sion and degradation equations or complex computer models. The
objective is to estimate the chemical concentration in the exposure
medium at the exposure point. Finally, human intake of or  contact
with the exposure medium must be modeled to allow projection of
the chemical dose.
Characterize Risk
  This step is the integration of the information developed in the
previous steps of the risk analysis. The risks of various adverse ef-
fects  are  calculated for each indicator chemical and each ap-
propriate exposure pathway by combining the projected doses with
the dose-response data. Various risk measures can be developed,
including  maximum and average individual  risk and  population
risk. The exact procedure for combining this information will be
prescribed  by the dose-response model selected and may vary for
different chemicals. For example, a linear non-threshold model
could be used to evaluate carcinogens, and non-carcinogens could
be  assessed based on a threshold response. Because of the  large
amount of risk information generated, it  may be desirable to ag-
gregate some risk estimates at this point. For example, risks for
pathways  affecting  the same population  could  possibly  be ag-
gregated, as could risks from all carcinogens.

APPLICATIONS FOR COMPUTERS
IN RISK ANALYSIS
   Given a risk analysis  approach such  as that described in the
previous section or  a similar quantitative approach, there  are two
ways to proceed. All the calculations and data management can be
done by hand or computers  can be used to make the analysis easier.
• For sites at which only a qualitative or simple quantitative assess-
ment is required, the  use  of computers would  be unnecessary.
However, for waste sites having many chemicals and several ex-
posure pathways, use of computers can both enhance and facilitate
the risk analysis.
  Computers can facilitate  a risk analysis in many ways.  Perhaps
most significantly, fully computerized risk models allow  a  more
thorough  treatment  of the  unavoidable uncertainty in any risk
analysis. Using a sensitivity  or bounding analysis, the reasonable
range of risks for a site can be estimated. Computers  also have ap-
plications  in the individual components  of the  risk analysis,
especially fate and transport modeling. Available computer models
account for more complex transport processes and use more site-
specific data than would be possible using desk-top calculations.
The computer models provide  more realistic predictions of en-
vironmental concentrations. A third significant way that computers
can assist a risk analysis is in data management. The large amounts
of input data needed  to conduct  a site  risk assessment  can  be
organized  in computer data bases. The intermediate and output
data produced during the analysis can be tracked by computer, and
statistical and graphics packages can be used to produce additional
outputs.
  Another, more limited, application of computers in risk analysis
is data gathering through telecommunication with centralized data
                                                       bases,  especially data bases of chemical properties. Examples of
                                                       useful on-line data bases are the Toxicology Data Bank (TDB) and
                                                       the Registry of Toxic Effects  of Chemical Substances (RTECS),
                                                       both maintained by the National Library of Medicine.
                                                         Brief  descriptions  of  two computer  applications  in  risk
                                                       analysis—data management by personal computer (PC) and use of
                                                       computerized transport models—are given below.

                                                       Data Management by PC
                                                         PC spreadsheet and data base management packages can be used
                                                       to manage the data generated during a site risk analysis. Obviously,
                                                       the risk  analysis approach  outlined in the second  section can
                                                       generate large amounts of data for even a moderately complex site.
                                                       Use of a PC to record, organize and store the data can  facilitate the
                                                       analysis.  For example, suppose that a site  risk  analysis was being
                                                       done for ten chemicals, six exposure pathways and four remedial
                                                       alternatives to  estimate chronic  and subchronic  risks.  In  this
                                                       hypothetical analysis, there could be as many as 480 (10 x 6 x  4  x
                                                       2) separate risk pathways to be quantified.  Maintaining these data
                                                       on a PC would greatly assist the  analysis. Examples  of PC data
                                                       tables,  created for a hypothetical risk analysis using the Lotus 1-2-3
                                                       spreadsheet package,  are given in Figures 2 and 3. The table  in
                                                       Figure  2 contains chemical concentration estimates  for various
                                                       population exposure  points. The table in Figure 3  gives  dose
                                                       estimates for  the same chemicals  and exposure points,  with the
                                                       algebraic conversion done by the PC. Given the dose-response fac-
                                                       tor for  a chemical, risk values also can be calculated and aggregated
                                                       on the  PC.
                                                       Computerized Fate and Transport Models
                                                         Many computer models have been developed to predict the fate
                                                       and transport of chemicals in the environment, and computer codes
                                                       for  many of these are available.  Several useful  models are  also
                                                       available through  the USEPA's  Graphical  Exposure  Modeling
                                                       System (GEMS),  which contains  transport models for ground-
                                                       water, surface water and air. GEMS is designed to be user-friendly
                                                       and its models can be accessed via telecommunications  with the
                                                       USEPA host computer. Examples of the fate and transport models
                                                       available in GEMS include:
                                                       •ATM—Gaussian dispersion model for predicting contaminant
                                                        fate and transport in air
                                                       •AT123D—three-dimensional model for predicting fate and trans-
                                                        port of contaminants in groundwater
                                                       •SESOIL—unsaturated zone transport model
                                                       •EXAMS—three-dimensional model for evaluating surface water
                                                        fate and transport of chemicals
                                                         Computer models generally give much more detailed  estimates of
                                                       environmental concentrations over time and space than are other-
                                                       wise possible. They can often  be used to construct maps of con-
                                                       taminant isopleths over time. This type  of  detailed environmental
                                                       fate information  allows greater  flexibility in  the risk  analysis.
                                                       However, these models only address one component of the  risk
                                                       analysis approach, and they must be integrated into the overall
                                                       analytical framework for the analysis.

                                                       INTEGRATED RISK ANALYSIS MODELS:
                                                       AN EXAMPLE

                                                         Another application of computers to  risk analysis is the in-
                                                       tegrated risk model, defined here as a model that has all or several
                                                       of the necessary components integrated into a single structure. An
                                                       example would be a model that estimated transport, exposure and
                                                       risk for a defined chemical release. In these models,  the algorithms
                                                       are linked so that risks can be  calculated for a  given set of inputs
                                                       without the need for intermediate calculations. The RCRA Risk-
                                                       Cost Analysis Model' would be one example of an  integrated risk
                                                       model, although this model is not  suitable to specific site analysis.
                                                         To illustrate the concept of  an integrated hazardous waste risk
                                                       model, a model being developed for hazardous waste land disposal
                                                       is described briefly in this  section. With a  few modifications, this
                                                       model  could be adapted for risk analysis at some types of hazar-
302
RISK ASSESSMENT/DECISION ANALYSIS

-------

ChMical Hue
Vinyl
Chloride




1,2-Dichloroethane




Alternative
A
B
C
D

A
B
C
D

Type
Han.
Avg.
Kax.
Avg.
Max.
Avg.
llax.
Avg.

Nan.
Avg.
hax.
Avg.
Hax.
Avg.
Hax.
Avg.
Ground
Public
Drinking Hater
3.1E-03
l.OE-03
3.1E-04
l.OE-04
O.OE+00
O.OE+00
1.6E-03
B.OE-04

4.1E-03
2.0E-03
6.1E-04
2.0E-04
O.OE+00
O.OE+00
3.1E-03
1.5E-03
Hater lig/1)
Private
Drinking Hater
6.BE-03
4.2E-03
6.BE-03
4.2E-03
5.7E-05
5.2E-05
5.7E-03
4.3E-03

1.3E-02
8.0E-03
1.3E-02
8.0E-03
1.1E-04
l.OE-04
1.1E-02
B.3E-03
Surface Hat«r

Hell head
Treatient
O.OE+00
O.OE+00
7.4E-01
2.1E-02
O.OE+00
O.OE+00
O.OE+00
O.OE+00

O.OE+00
O.OE+00
1.4E+00
4.0E-02
O.OE+00
O.OE+00
O.OE+00
O.OE+00

Abateient
Plant
O.OE+00
O.OE+00
O.OE+00
O.OE+00
1.2Łt(W
3.4E-02
5.1E+00
3.0E-02

O.OE+00
O.OE+00
O.OE+00
O.OE+00
2.SE+00
6.5E-02
9.9E+00
5. BE -02
                                                            Figure 2
                                         Sample Data Management Format for Risk Analysis:
                                                     Chemical Concentrations
dous waste sites. The primary modifications required for this ap-
plication would be in the chemical release module and in the site-
specific hydrogeologic data required. The model would be most ap-
plicable to sites where ground water is the major exposure pathway.
In any case,  the  model's structured, logical approach to risk
analysis could be applied.
Structure of the Model
  The basic risk analysis approach taken in the model is similar to
that described in the second section. The model is a computerized
set of algorithms that projects chronic health risks over a specified
time horizon for land disposal facilities. The model has four major
modules  linked into a continuous structure.
  The  failure/release module determines  the year of failure (de-
fined as release of leachate to groundwater) for a facility based on
its design and location. It also estimates the annual volume of
leachate released to groundwater based on a facility's design, size
and climate category (defined by net infiltration). At present, the
model  contains several landfill  and surface impoundment design
types. Clay and synthetic liners in various single- and double-layer
design  configurations plus unlined facility  designs are included.
The current release module would be inapplicable to most uncon-
trolled  waste sites, with the possible exception of sites with aban-
doned  landfills or surface impoundments.  An alternate release
module would most likely have  to be incorporated into the model
for application at uncontrolled waste sites.
  The groundwater transport module estimates  chemical concen-
trations in groundwater and chemical loadings to surface water
resulting from leachate releases by a facility. The groundwater
transport module produces time profiles of concentration and mass
loading at various down-gradient distances  from  the  source.
Chemical-specific mobility and persistence in groundwater are fac-
tored into the analysis. Several groundwater flow scenarios, in-
cluding single- and multi-layer aquifers with a range of horizontal
and vertical velocity vectors, are contained in the model. Waste site
applications would  need to incorporate certain site-specific data
such as porosity, dispersivity, groundwater velocity and direction,
unsaturated zone characteristics and aquifer thickness.
  The human exposure module  estimates the lifetime  dose of
chemical received through either groundwater or surface  water
sources of drinking water. A time profile of lifetime dose is derived
from the chemical concentration profile estimated previously. The
resulting dose profile is the quantitative representation of chemical
exposure used to estimate risk.
  The dose-response module calculates lifetime risks of various
chronic health effects for the estimated chemical doses. Risks are
estimated separately for each chemical constituent of the waste. A
time profile of individual risk is produced by the module,  and
population risk is estimated by multiplying the individual risk pro-
file by the population profile.
  The four  modules described above are linked by appropriate
algorithms to form  the basic structure of the risk model.  In addi-
tion,  the computerized model includes several input data bases
from which it retrieves information and input and output  modules
that facilitate operation of the model and presentation of the
results.

Operation of the Model

  The model requires certain input variables to be specified by the
user.  The input  variables include chemical (or  waste)  identity,
                                                                         RISK ASSESSMENT/DECISION ANALYSIS
                                                          303

-------
climate category, groundwater flow scenario, distance from facility
to exposure  point  and  population  exposed. Additional  inputs
would be required for site-specific applications at waste sites.
  The calculation sequence for a typical model analysis is shown in
Figure 4.  In an actual model  run, this sequence is repeated as
directed by the program until risk estimates are  calculated  for all
specified variable combinations. In addition to  the operating se-
quence, this  flow chart shows  the input variables and chemical-
specific parameters that are factors in the major  calculations.
  The first step in the model sequence is determination of the year
of facility failure and  the leachate release  rate following failure.
The model then estimates the concentration  of specific chemicals in
the leachate and calculates a chemical release rate. The next key
step is calculation of the time delay resulting from chemical passage
through the  unsaturated  groundwater zone. At this  point, the
model has estimated the total time for the chemical to reach the
saturated  zone and the annual  mass input  of the chemical  to the
saturated  zone.
  The next step is estimation of groundwater chemical concentra-
tion at  the specified exposure point  over the time horizon. This
computation  is  based  on the  groundwater flow  scenario, the
distance to the exposure point,  the mobility and persistence of the
chemical  being  evaluated and the  chemical   mass input  rate
previously calculated. Following this, the model calculates lifetime
chemical doses. The individual risk is estimated on the basis of the
dose  and  chemical-specific toxicity parameters.  The final  major
step is estimation of the population risk. At this point, all  of the
basic output variables have been calculated. They can then be sum-
marized, statistically evaluated  and displayed in  a variety of ways
depending on the specific objectives of the  analysis.
                                                        Applicability to Uncontrolled Hazardou§ Waste Sites
                                                          The illustrative model described above must be adapted for ap-
                                                        plication to uncontrolled waste sites. However, with some adjust-
                                                        ment, it contains all of the basic components necessary to estimate
                                                        risk for a waste site. More significantly, the model has a flexible
                                                        structure that can  be adapted to  different hazardous waste situa-
                                                        tions. The model is clearly most applicable to sites where ground-
                                                        water contamination is the primary route of potential human ex-
                                                        posure.
                                                          To  adapt this model to an  uncontrolled hazardous waste site,
                                                        chemical releases from the site must be estimated. This is typically
                                                        one of the most difficult problems in assessing risk at uncontrolled
                                                        waste sites.  Because the model is fully computerized, a range of
                                                        release estimates can be rapidly analyzed and the sensitivity of risk
                                                        to various release profiles evaluated. The model is well suited to
                                                        sensitivity   analysis  when   dealing  with  the  unavoidable
                                                        methodologic and data uncertainties.
                                                          Certain site-specific data would also  be required  for application
                                                        of the model to a site. At a minimum, basic information about site
                                                        geohydrology  and  meteorology   and  analytical   data  on the
                                                        chemicals present  must be available. The degree of uncertainty in
                                                        the resulting risk estimates will  in  part  be due to the quantity and
                                                        quality of site-specific data available. Again, sensitivity analysis can
                                                        be used to evaluate the effects of assumptions used to fill data gaps.

                                                        CONCLUSIONS
                                                          Systematic risk  analysis can  assist rational decision-making at
                                                        uncontrolled hazardous waste sites by providing information on
                                                        projected health effects resulting from chemical releases at the site.


Nate Data
Vin,I
Chloride




1,2-Dichloroetliane





Alternative
A
B
C
D

A
B
C
D


Type
flat.
Avg.
flai.
Avg.
nit,
Avg.
Na«.
Avg.

Nai.
fla«.
Avg.
Hai.
Avg.
Hat.
Avg.
Ground
Itg /
0 k 1
Drinking Nater
9.5E-05
3.IE-05
9.5E-04
3.IE-04
O.OE'OO
O.OE'OO
4.9E-05
2.5E-05

I.9E-04
4.2E-05
I.9E-U5
4.TE-04
O.nEtOu
U.OE'OO
9.5E-05
4.4E-05
Hater
kg/day)
Or »«
Drinking Niter
2.1E-04
I.3E-04
2.1E-04
I.3E-04
l.BE-04
I.4E-04
I.8E-04
1 . 3E-04

4.0E-04
2.5E-04
4.0E-04
2.5E-04
3.4E-04
3.1E-04
3.4E-04
2.4E-04
Surface Hater
(Recreational Creek I
(ig/kq /day)

SKII Intake Fish Intake
O.OE'OO O.OE'OO
O.OE'OO O.OE'OO
O.OE'OO O.OE'OO
O.OE'OO O.OE'OO
6.0E-04 9.1E-07
8.7E-09 3.4E-08
2.8E-05 4.3E-04
8.1E-09 3.2E-08

O.OE'OO O.OE'OO
O.OE'OO O.OE'OO
O.OE'OO O.OE'OO
O.OE'OO O.OE'OO
I.2E-OS I.6E-04
1.7E-08 6.4E-08
5.5E-05 8.4E-04
I.4E-OB 4.4E-08

a
Treatunt Plant
3.5E-02
6.5E-04
J.tt-03
4.5E-'5
O.OE'OO
O.OE'OO
1.4E-02
4.8E-04

4.7E-02
1.2E-03
4.7E-OJ
1.2E-04
O.OE'OO
O.OE'OO
2.7E-02
9.2E-04
Air
,
TreatMflt
0.0ٻ00
O.OE'OO
I.7E-OI
4.8E-03
O.OE'OO
O.OE'OO
O.OE'OO
O.OE'OO

O.OE'OO
O.OE'OO
3.2E-01
9.2E-03
O.OE'OO
O.OE'OO
O.uE'OO
O.OE'OO


Plant
O.OE'OO
O.OE'OO
O.OE'OO
O.Of'OO
2.8E-01
7.8E-03
1.2E'0«
4.9E-03

O.OE'OO
O.OE'OO
O.OE'OO
O.OE'OO
5.8E-OI
1.5E-02
?.3E»00
I.3E-02
Total Eiposure
l*f/ k»/dayl

Public Private
9.5E-05 3.H-02
J.IE-05 7.7EHM
9.5E-04 1.7E-41
3.1E-04 5.0E-03
4.9E-04 2.K-01
4.3E-08 7.BHJ
8.2E-05 1.2E««0
2.5E-05 7.K-03

1.9E-04 6.7E-02
4.2E-05 1.5E-OJ
I.9E-05 3.2E-01
4.TE-04 9.K-OJ
I.4E-OS 5.8E-01
B.JE-48 I.5E-02
1.4E-04 2.JE'00
4.4E-03 I.4E-02
                                                            Figure 3
                                   Sample Data Management Format for Risk Analysis: Chemical Doses
 304
RISK ASSESSMENT/DECISION ANALYSIS

-------
Design Type
Climate
Facility Size
Constituent Mass
Design Type
Facility Size
Unsaturated Zone
  Thickness
GW Flow Scenario
Distance      —
Facility Size
Determine Year of Failure
and Leachate Release Rate
                       Calculate Constituent
                           Release Rate
       Calculate
 Unsaturated Zone Delay
                                                    Solubility
Mobility
Solubility
 Calculate Groundwater
 Concentration Profile
                        Calculate Lifetime
                          Dose Profile
                            Estin-ate
                      Individual Risk  Profile
                                                     Toxicity
Populat ion
                            Estimate
                      Population Risk  Profile
                            Figure 4
                 Risk Analysis Model Flow Chart
Risk analysis can be used to determine the need for action at a site,
to assist in selecting and designing a remedial action for a site and
to evaluate the effectiveness of actions taken at a site. Risk analysis
can also enlighten certain public policy questions, such as setting
site cleanup priorities and evaluating overall program effectiveness.
  The structured risk analysis approach outlined in this paper can
provide not only numerical estimates of risk but, more importantly,
can also serve as the organizational framework for a large amount
of  site-specific data  related  to  potential health  effects.  Risk
analysis, however, is not without its limitations, including inherent
methodologic uncertainties and the requirement for considerable
input data.
  In a number of ways, computers can facilitate the risk analysis
process. Computers can be used to manage the risk  analysis  data.
Off-the-shelf computer   models  for  environmental  fate  and
transport of chemicals can be used to assist in the exposure evalua-
tion, and integrated computer risk analysis models, such as the il-
lustrative model described in the previous section can be developed
and used at the site.
  An integrated computer modeling approach generally permits
greater breadth and depth in the analysis by reducing computation-
time and allowing rapid  sorting and manipulation of the  data.
More complex computation  algorithms can be  used,  and  more
combinations of variables can be analyzed. The integrated  com-
puter model also allows a more thorough treatment of uncertainty
by  facilitating rapid sensitivity analysis. A disadvantage of  in-
tegrated computer  models is the front-end time and cost required
for development.  This generally  makes it impractical to develop
and use a fully integrated model for a single site. By developing a
generic  model that can be applied with slight modification to a
group of sites, or by adapting existing models, this disadvantage
can be overcome.

REFERENCES
1. National Academy of Science, Risk Assessment in the Federal Gov-
  ernment, National Academy Press, Washington, DC, 1983.
2. Ruckelshaus, W., "Science, Risk, and  Public Policy," Science 221,
   1983, 1026-1028.
3. USEP A, in Health Assessment Document for Polychlorinated Dibenzo-
  p-Dioxins. Review Draft. EPA-600/8-84-014A. Environmental  Criteria
  and Assessment Office, Cincinnati, OH, 1984, Table 11-36.
4. USEPA, Health  Effects Assessments for  various  chemicals. Draft.
  Environmental Criteria and Assessment Office, Cincinnati, OH,  1984.
5. ICF,  The RCRA  Risk-Cost Analysis Model. Phase HI Report; pre-
  pared for the Office of Solid Waste, USEPA, Washington, DC,  1984.
                                                                          RISK ASSESSMENT/DECISION ANALYSIS
                                                                                                          305

-------
           COPING WITH  UNCERTAINTY  IN EVALUATING
                      ALTERNATIVE REMEDIAL ACTIONS

                                       WILLIAM A. TUCKER, Ph.D.
                                        GREGORY J. GENSHEIMER
                                           ROBERT F. DICKINSON
                                Environmental Science and Engineering,  Inc.
                                               Gainesville, Florida
INTRODUCTION
  Remedial investigation/feasibility studies (Rl/FS) typically de-
mand competent decision-making in the face of extreme uncertain-
ty. Significant sources of uncertainty include: (1) representativeness
of soil and groundwater samples taken from grossly heterogeneous
media, (2) limited availability of data at critical decision points and
(3) inadequate theoretical basis for evaluating contaminant fate and
transport. In this paper, the authors discuss uncertainties presented
by the above sources. Techniques presented in this paper are useful
at several steps in the RI/FS process, including: (1) development of
response objectives and evaluation criteria, (2) design of field sam-
pling strategies and (3) evaluating costs and effectiveness of alter-
native actions.
  Several case studies are  presented. Although each case study is
based on experiences of the senior author in evaluating remedial ac-
tions at contaminated sites, the examples have been  simplified
slightly to clearly illustrate concepts. The case studies include:
•Estimating the volume of contaminated sediments and soils in the
 drainageway of a large hazardous waste landfill. This study il-
 lustrates the utility of Monte Carlo-like procedures in estimating
 the probable depth of penetration of strongly sorbing organic
 chemicals and the value which can be derived from limited avail-
 able data.
•Applying statistical techniques to the problem of estimating the
 volume of pumpage required to restore an aquifer contaminated
 by mixed solvents; the potential risks associated with using "best
 estimates" are demonstrated by this case study.
•Utilizing (and thus illustrating the value of) kriging in estimating
 the volume of  contaminated soils and in developing sampling
 strategies.

CONTAMINATED SEDIMENT VOLUME ESTIMATION

  A small stream received runoff from a large hazardous waste
landfill which began receiving wastes in 1962. Adjacent to the land-
fill is a small industrial facility which also discharged cooling water,
obtained  from  an on-site well, to the stream. The  industrial
discharge maintains a fairly steady flow in the stream which, under
natural  conditions, would  be intermittent or ephemeral.
  The natural groundwater elevation is 18 ft below the stream bed
elevation. Thus, the stream is  an artificial  recharge zone from
which a steady flow of water percolates to the water table at a rate
determined by the permeability of the underlying soils. The most
toxic contaminants associated  with  the  site are chlorinated
hydrocarbons with very high soil adsorption coefficients. Potential
                                                   remedial alternatives are: (1) stream diversion and capping, (2) ex-
                                                   cavation and disposal in the landfill and (3) groundwater restora-
                                                   tion coupled with either of the first two steps.
                                                     Critical to the evaluation of these alternatives are:
                                                   •Potential for groundwater contamination; i.e., have contaminants
                                                    migrated to the water  table
                                                   •Estimation  of the volume of contaminated materials—a large
                                                    volume argues against excavation based on cost, health and en-
                                                    vironmental risks. The surficial contamination of the bed sedi-
                                                    ments has been mapped, so the key unknown is the depth of con-
                                                    taminant penetration.

                                                   Contaminant Penetration

                                                     The  following  equation describes the depth of contaminant
                                                   penetration:
                                                                              TKV
                                            (1)
                                                     where

                                                       T  =
                                                     Kv   =
                                                       n  =
                                                     pb   =
                                                    1C,   =
                                                      loc
                                                       L
            n +  pb KOC foe

time since wastes were first stored at the landfill
vertical saturated conductivity
soil porosity
soil bulk density
organic carbon adsorption coefficient
fractional organic carbon content of the soil
depth of contaminant penetration
                                                     With the exception of the facility lifetime, each of the parameters
                                                   of Equation 1 must be estimated. Conductivity has been measured
                                                   in similar soils associated with the landfill, as well as other sites in
                                                   the general area. These data appear to be log-normally distributed
                                                   with a median value of 10.4 ft/yr and a log-standard deviation of
                                                   0.45. In other words, the estimated permeability is correct to about
                                                   a factor of 4.
                                                     Virtually no information is available on the organic carbon con-
                                                   tent of the soils. County soil surveys, related USDA data and soils
                                                   texts such as Brady2 indicate that soils of the type found here may
                                                   range from  0.1 to 3%  organic carbon.  The porosity also is
                                                   unknown, and standard texts' indicate that it may range from 0.25
                                                   to 0.5
                                                     The organic carbon adsorption coefficient of the most toxic con-
                                                   taminants known  to be present has been estimated  to be 10*.
                                                   Methods and data presented by Lyman, el al.' indicate that the er-
                                                   ror associated with this estimate is about a factor of 8.
306
RISK ASSESSMENT/DECISION ANALYSIS

-------
Monte Carlo Simulation
  Two techniques which can rationally account for the uncertainty
are compared. The first is the classic Monte Carlo procedure. The
available data indicate  that uncertainties in Koc and Kv are log-
normally distributed. The uncertainties in n and foc are assumed to
be uniform through their full range since there is no basis for even
making a "best" estimate. The Monte Carlo analysis was per-
formed readily and at low cost using the Statistical Analysis System
(SAS) on the University of Florida computer system (programming
time 1  hr, computer cost $5 for N = 10,000). The results may be
summarized as follows:
    The median penetration depth is  only 0.8 ft., but there is a
    7%  chance  that contamination has migrated  18  ft to  the
    water table.
Sensitivity Analysis
   An alternative method of analysis drawing on the Monte  Carlo
philosophy, but which can be performed quickly by hand calcula-
tion, was also developed.  This method  short circuits  the Monte
Carlo  procedure by  sampling  the  distribution  of  the  input
parameters at fixed, predetermined  points. This method was ex-
pected to reproduce the central tendencies of the Monte  Carlo
distribution,  recognizing that the "tails" of the distribution (ex-
tremely unlikely combinations) would not be accurately repro-
duced.
   The proposed  method is most appropriately characterized as a
sensitivity analysis  which  realistically incorporates  the  expected
uncertainty distributions of the input  parameters. The method is
exemplified by looking at the distribution function assumed for Koc
(Fig. 1). In Monte Carlo analysis, this  distribution is sampled ran-
domly for input to the calculation of penetration depth. A large
random sample is required to represent the distribution accurately.
   In the alternative method, the distribution is sampled at predeter-
mined percentiles, each of which represents an equal fraction of
the complete distribution. For example, if four  "samples"  are
selected they would be  12.5, 37.5, 62.5 and 87.5 percentiles. Each
sample realistically represents 25% of the range. If only three were
selected, they would be the 17, 50 and 83 percentiles. In the latter
case, each sample represents 33% of the range. Each of the sam-
pled Kocs represents a range of values with equal probability  of oc-
currence.  Each  uncertain input is sampled by the same strategy,
and all possible combinations  of inputs are passed through the
equation. For this analysis, three points were selected from each
distribution, resulting in 81 (34) combinations of parameters and 81
independent and equally likely estimates of the penetration depth.
  An advantage of this method is that fewer independent estimates
of penetration depth must be calculated, thus making the analysis
possible by hand and without need of a table of random numbers.
The question is whether the method yields results  similar to the
classical  Monte Carlo result. The results, which are compared in
Figure 2, show the uncertainty  distribution  of the calculated
penetration  depth.  It is apparent that  the "short cut" method
reproduces the central  tendency (from 25 to 75 percentile) of the
distribution accurately,  but is less accurate for the less likely ex-
tremes. An alternative analysis by the short cut method, but using
256 (44) equally likely combinations of inputs, is  not shown, that
analysis showed an improvement in the 90 to 95  percentile range
when compared with the 81 (34) calculation test.


Method Accuracy
  The accuracy of the distribution is the kind of test which might
interest a statistician. From a practical standpoint, though, one is
interested in whether the simple method provides information ac-
curate enough to support the right remedial action decision. By this
test,  the  short cut  method serves well. The short cut method in-
dicates the odds that contamination has moved to the water table
are 4%, while the Monte Carlo analysis indicates that the odds are
8%. Sound remedial action decisions should be insensitive to these
small differences in probability. Both methods indicate the most
likely depth of penetration is 0.8 ft.
  Suppose, next, that a few limited soil borings were taken by hand
coring methods to a depth of 4 ft and analyzed for site-related con-
taminants. Suppose, also, that  each of these  borings was  con-
taminated through its full depth. How can this new information be
used? Clearly, contamination has penetrated deeper than the best
estimate; however, one must remember that the analysis indicated a
20% change of penetration below 4 ft, so it is not all that discon-
certing to find contamination  that deep. This new information
shows that  some combinations of the independently estimated
values are not possible.  All combinations which yield penetration
depths of less than 4 ft are  clearly impossible and should be deleted.
                S  10      25      SO      75     M  »5     W
              PERCENT PROBABILITY THAT Koc IS LESS THAN INDICATED VALUE
                           Figure 1
       Uncertainty Distribution for the Absorption Coefficient
                                                                  WATER TABLE
           1      6 10     25     SO     7S     W  tS     M
     PERCENT CHANCE CONTAMINATION HAS PENETRATED LESS THAN INDICATED DEPTH


                            Figure 2
           Uncertainty Distribution for Penetration Depth
                                                                         RISK ASSESSMENT/DECISION ANALYSIS
                                                                                                                            307

-------
      i.oooT---t
                   I  I
WATER TABLE
            <      I 10      »      M     n    to   it     n
      PERCENT CHANCE CONTAMINATION HAS PENETRATED LESS THAN INDICATED DEPTH


                           Figure 3
   Uncertainly Distribution for Penetration Depth after Soil Analysis
   These procedures were  reapplied under  this constraint. The
uncertainty  distributions calculated for penetration depth by the
two procedures are shown in Figure 3. Now the best estimate depth
by Monte Carlo is 11.6  ft, while the short cut result is 9.4 ft. The
odds that contamination has reached the groundwater are 34% by
Monte Carlo and 14% by the short cut. It is  interesting that, after
learning that contamination has  already  penetrated  4 ft,  the
statistical methods indicate that it is more probable that the con-
tamination  has penetrated to  the  water  table.  The methods
presented here can use this  data effectively.
   Decisions based on "best estimates" would necessarily be con-
fused by data in contradiction with the best  estimate. Use of best
estimates as inputs will lead to an estimate of penetration depth of
less than 1 ft. When samples showed contamination as deep as 4 ft
there would be a tendency to lose faith in such transport estimates,
although they would clearly be needed to complete the assessment.
   Methods based on "worst case"  could not really use  the new
data in any meaningful way. The "worst case" estimate is that con-
taminant has migrated to the water table. Confirmation of penetra-
tion to a depth of 4 ft would have no influence over the decision
process if "worst case"  estimates were relied on.

AQUIFER RESTORATION

   A spill of mixed  solvents resulted in a plume of contaminated
groundwater in a shallow sand aquifer. The preferred remedial ac-
tion is installation of three groundwater extraction wells. The con-
taminated groundwater will be discharged to a sanitary sewer and
ultimately will be treated in a POTW that has  secondary treatment.
   To properly evaluate the cost of this alternative, it is essential to
estimate the volume of water which must be pumped and the dura-
tion of the pumping required before the aquifer water quality will
meet acceptable water quality standards. For simplicity, this discus-
sion focuses on benzene,  one of the several solvents which currently
exceeds those standards.
   Measured benzene levels in the contaminated area ranged from 1
to 5 /tg/1. Four sampling  locations were so  heavily contaminated by
other solvents, however,  that high concentrations of benzene could
have been masked by interference. In  one of these locations,  the
detection limit for benzene  was  500 pg/1. The average concentra-
tion of benzene in the contaminated area could  be as high as 200
ftg/1 if an estimation is made using conservative  assumptions.
                                                           A  groundwater  flow model was  used to design an  efficient
                                                         shallow pumping well network to draw uncontarninated water from
                                                         outside the contaminated zone through the contaminated area. One
                                                         pore volume of clean  water would  be drawn  through  the con-
                                                         taminated zone every month with the wells in operation. The ques-
                                                         tion posed was:  how long must the  groundwater extraction con-
                                                         tinue before  benzene levels would be reduced to 1 ^g/1?
                                                           The concentration in the groundwater after N pore volumes of
                                                         water have been extracted, Cn, is given  by:
                                                                                       SL- ('
                                                                                       C,   \
                                                                             !*-•  ('-	"—^
                                                                             ct/l  \    n t Ptk«f«y
                                                                                  \       <-»««/



                                                                                  ('-»*  ".K,r,)
                                                                                                                             (2)
                                                                                 log 0.01 - N l»|
                                                        where C0 is the initial concentration.
                                                           The bulk density, porosity and organic carbon content of the soil
                                                         (^b  = 1.9, n  = 0.35 and f^ =  0.002, respectively) were all deter-
                                                         mined via  sampling and soil characterization. The variability evi-
                                                         dent in these  data could also be used in a Monte Carlo analysis
                                                         similar to the  one used in the first example, but for the sake of il-
                                                         lustration, it will be ignored in this discussion. The result is sensitive
                                                         to uncertainty in the K^ value for benzene. Benzene's K^ has been
                                                         measured by Karickhoff, el al.' so the uncertainty is considerably
                                                         less  than usual for estimated values. Still, Karickhoff, el al.' data
                                                         suggest that measured K^ values are only reliable  to about 20%.
                                                           The uncertainty is assumed to be log-normally distributed about
                                                         a  median value of 83 with a logarithmic standard deviation of 0.1.
                                                         This condition may be  compared with  the logarithmic standard
                                                         deviation of 0.78 which  was used tor estimated K^. values in  the
                                                         previous case study. Using the best estimate value:
                                                                            .jOOl
                                                                lo,
                                                                          0.3! - 1.9 x 5!
                                                                                                                   (3)
                                                         so 6 pore volumes must be drawn through the contaminated zone to
                                                         achieve the cleanup standard. This will take 6 months of pumping.
                                                         Given the uncertainty in K^, however, there is a 50% chance that
                                                         the standard will not be achieved in 6 months. What is the "upper
                                                         bound" on the required pumping time? Applying the log-normal
                                                         uncertainty,  one can say that one is 95% confident that the actual
                                                         Koc for benzene is less than  121. Then one can be 95% confident
                                                         the duration of pumping required to reach standard is less than 8
                                                         months. The restoration of the aquifer will  almost certainly be
                                                         achieved in less than 1 year.
                                                         UTILIZATION  OF KRIGING

                                                           Surficial soils at an abandoned industrial facility have been con-
                                                         taminated  by a  variety of aromatic hydrocarbons.  Groundwater
                                                         underlying these soils is not contaminated.  For the sake of Illustra-
                                                         tion, this discussion will focus on pyrene, one of several identified
                                                         contaminants. Again, strictly for the purposes of this presentation,
                                                         the  authors  will assume that  the affected parties  have  reached
                                                         agreement with regulatory  authorities  that the  soil's  cleanup
                                                         criterion will be 100 ppm, based on the potential for human ex-
                                                         posure via drinking water.
                                                           Surficial soils were collected at locations  shown in Figure 4.
                                                         Thirty-seven samples were taken including one 14-point  transect
                                                         along the long axis of the site, intersected at three points by smaller
                                                         perpendicular transects.  Distances between points in the transects
                                                         were irregular.  Seven  other  points were placed randomly in  the
                                                         field, but nearer the intersections of the transects. At  each point,
                                                         surficial soil  samples were analyzed for the contaminant concentra-
                                                         tion.
308
RISK ASSESSMENT/DECISION ANALYSIS

-------
  Because environmental quality indicators often tend toward a
log-normal  distribution,  contamination  levels  were  log-trans-
formed prior to statistical analysis. After log-transformation, most
statistical procedures based on  an assumption of normality are
directly applicable.
  The pattern of contamination is quite sporadic, with little spatial
coherence.  One  location exceeds 1,000 ppm; another location ex-
ceeds 100 ppm.  These sampling locations were not adjacent, and
samples taken between themshow  degrees of contamination vary-
ing from undetectable to 40 ppm. Such heterogeneity is common in
soil sampling.  Since soil contamination is often highly variable over
short inter-sample distances, routine interpolation techniques can
be misleading, implying considerable confidence in the concentra-
tion values at unsampled locations. To avoid such inappropriate in-
terpretations,  Environmental Science and Engineering, Inc. (ESE)
has recently applied for geostatistical approach known as kriging
on  several assignments of this  type.
Kriging Theory

  Kriging is a method of interpolation  based on the theory  of
regionalized variables. Regionalized variables are those variables
that vary geographically, such as rainfall, metal content of an ore
body or, in this  application, soil contamination levels. The advan-
tage of Kriging  over other interpolation methods is  that it is op-
timum  and unbiased.  The greatest  benefit is the  fact that  it
calculates an error for each estimated point.
  Kriging makes use of the semi-variogram, a graph of the average
differences (semi-variance) between pairs of points  at increasing
separation  distances. Theoretically,  the  semi-variance increases
with separation distance from  0 to a value equal to the population
variance. The inter-sample distance  at which the semi-variance
        3
    E
a  a     a    •
    a
    a
    a
a  E    g  E
     A
    a
 *  a    a    a
    a
     a
     a
     KEY
      D   CONTAMINATION PRESENT
      A   SHALLOW SOIL SAMPLE
                                  levels off at the value of the population variance is called the range.
                                  Inside the range, pairs of points are somehow correlated; outside
                                  the  range, they are independent. Kriging is best used when the
                                  distances between the  points to be estimated and the measured
                                  values are shorter than the range.
                                    Kriging was used to contour map the boundary between con-
                                  taminated and uncontaminated soils. Since all soils contaminated
                                  above standard might have to be excavated, realistic estimates of
                                  the area had significant cost implications. Kriging was used to max-
                                  imize the investigators' certainty of the contour lines to minimize
                                  the costly chances of excavating uncontaminated soils.

                                  Semi-Variogram Estimation

                                    The semi-variogram estimation and universal  Kriging program5
                                  was used in this study. It consists of several options enabling the
                                  user to calculate semi-variograms, test goodness-of-fit (GOF) of the
                                  calculated semi-variograms to the population and estimate point or
                                  area values.
                                    The semi-variogram was calculated, and a smooth function was
                                  fit using a spherical model. Without going into further detail (see
                                  the article by Gambolati and Valpi'), the parameters used to fit the
                                  semi-variogram implied that: (1) points sampled  closer together
                                  than 200 m were correlated; (2) the variance of the population was
                                  approximately 1.26  log units; (3)  38% of the  variance was at-
                                  tributable to spatial variations; and (4) 62% of the variance was
                                  either random or could not be accounted for by the sampling
                                  scheme. The latter part of the variance is usually called the nugget
                                  variance or nugget.
                                                                        too
                                                                                        too
                                                                                10.100 ppm
                                                                                  MQ ppm
                                                                                  < 1 ppm
                            Figure 4
                         Sampling Grid
                                                             Figure 5
                                           Estimated Soil Contaminant Distribution by Kriging
                                                                          RISK ASSESSMENT/DECISION ANALYSIS       309

-------
  GOF tests involved the sequential removal of each data point,
replacement of that point by Kriging using the  semi-variogram
equation and neighboring points, and comparison of real versus
estimated points. A good fit was indicated if the average difference
between Kriged and real points was near 0 and if the reduced mean
square error (mean square error divided by the  variance of the
estimate) approached one. In this  study, the average error was
 -0.013, and the reduced mean square error was 0.964.
  About every 17 m 396 points were Kriged (interpolated) in the x
direction and 23 m in y. The x-y locations and Kriged z's were input
to Surface  II, a computer graphics package for contour mapping.
A very simple map was drawn (Fig. 5) using contours 1,10 and 100
ppm.
  Even though two data points had values greater than 100 ppm,
 the Kriging system did not reproduce them. The Kriging system
 regarded these two values as being anomalous, indicated by the
 large value of the nugget variance.  When the nugget is large, the
 Kriging  system tends to  smooth  the data. In this case,  the
 smoothing was excessive when compared to the actual data.
  To avoid this excessive smoothing, a new data set was:  (1)
 created by including the measured  values as well as the Kriged
 values and (2) used for a new contour map (Fig.  6). Areas of the
 map with values less than 1 ppm and values between 10 and 100
                                                       ppm were enlarged, and an area of values greater than 1000 ppm
                                                       was defined.

                                                       Uncertainties

                                                         The differences between Figures 5 and 6 indicate an uncertainty
                                                       in the placement of the contour lines. The error estimates at each
                                                       interpolated  point  were  contoured  in  Figure  7.  The  errors
                                                       represented one standard deviation for each estimate. The contours
                                                       genrally indicated that error increased with distance away from the
                                                       central location of the sample points. The greatest errors  were
                                                       located in the corners of the map  where samples had not  been
                                                       taken.
                                                         The best estimate and error estimate can be combined to estimate
                                                       the upper bound of the contaminated zone. The upper bound of
                                                       the contaminant levels at 90% confidence is shown in Figure 8. The
                                                       values were determined by adding 1.28 times the standard deviation
                                                       to the best estimate. A logarithmic standard deviation is 0.47, in-
                                                       dicated by the variance of replicate soil samples, was used for ac-
                                                       tual sample data points. The area with values greater than 100 ppm
                                                       grew from about 0.1 acres in  Figure 6 to about 6 acres.
                                                         Recognizing that soils exceeding 100 ppm must be addressed via a
                                                       remedial plan, the comparison of Figures 6 and 8 indicates the high
                                                       level of uncertainty. The high level of uncertainty also indicates a
  100
  400
  too
  eoo
                             G   X
                                                                    coo
                     100
                                       tOO
     KEY
     Ega 100-1,000 ppm
     tZfl   10-100 ppm
     E3     1-10 ppm
     I  I     < 1 ppm

                           Figure 6
      Best Estimate of Contaminant Distribution Based on Kriged
                   Estimates and Actual Data
                                                                    :oo
                                                                    zoo
                                                                    100
                                                               KEY
                                                           ESI UNCERTAINTY
                                                                LESS THAN
                                                                FACTOR OP 10
                                                           O UNCERTAINTY
                                                                GREATER THAN
                                                                FACTOR OF 10
                                                                                  Figure 7
                                                                  Distribution of Errors Associated with Kriging
 310
RISK ASSESSMENT/DECISION ANALYSIS

-------
need to further sample to either redefine the 100 ppm contour or
just to increase the certainty of estimates about that contour.
  The clearest way to illustrate the extent of uncertainty regarding
whether soils are above the cleanup criterion is to plot (Fig. 9) the
number of standard deviations between the best estimate value and
the criterion (100 ppm). For example, if the best estimate is 5 ppm
and the uncertainty in that  number is a factor of 5, one is still
relatively confident that the  soil does  not require cleanup. If the
uncertainty were a factor of 50, then it would be hard to guess
whether that soil would require action.
  In log units, the first example has a best estimate of  0.7 and a
standard deviation of 0.7, the latter has a standard deviation of 1.7.
Since the log of the criterion (100 ppm) is 2, the  estimate is 1.9 stan-
dard deviations below the criterion in the first case.  In the latter
case, the estimate is 0.8 standard deviations from the criterion.
  The parameter

  C =  [log (criterion) - log (estimate)]/ logarithmic standard deviation (3)

is directly related  to the confidence that  the soils are below
criterion. Large negative values (C <- 1.28) indicate soil contamina-
tion levels significantly above the criterion. High positive values
(C >1.28) indicate no further requirement for sampling because soil
contamination levels are significantly below the criterion.
  Areas for which the absolute value of C is less than 0.68 are areas
where there is very little basis for guessing  whether the soils are
above or below the criterion. In areas where the absolute value of C
is greater  than 0.68 but less than 1.28, there is a better than 75%
but less than 90% change  of  guesins whether the soils are within
standard.  Thus, areas where  ICI  s
                                  o
                      100

                                        206
                                                          300
      KEY
      BBI   > 1,000 ppm
      ES3 100-1.000 ppm
      i...:.3   10-100 ppm
      \'.'.:'.'.J     1-10 ppm
                            Figure 8
  Conservative Estimate of Contaminated Distribution; 90% Confidence
                                                                      soo- -^
                                                                      400
                                                                                                          ?00
        390% CONFIDENCE ABOVE STD
        } 75% CONFIDENCE ABOVE STD
        ] UNCERTAIN
        | 75% CONFIDENCE BELOW STD
        3 90% CONFIDENCE BELOW STD
        ] 95% CONFIDENCE BELOW STD
        I 99% CONFIDENCE BELOW STO
         RECOMMENDED ADDITIONAL
         SOIL SAMPLING LOCATIONS
                           Figure 9
     Confidence Zones with Respect to Soil Cleanup Criterion and
      Recommended Sampling Locations to Improve Confidence
                                                                           RISK ASSESSMENT/DECISION ANALYSIS
                                                           311

-------
have reached an aquifer. Adaptation of these procedures to dif-
ferent  amounts  and   qualities  of environmental  data  was
demonstrated in both that  case study and  the second  case study
which evaluated the probable duration of pumping required to
restore groundwater quality. The results are directly  applicable to
remedial action decisions.
   In a third case study, the geostatistical procedure known as krig-
ing was used to demonstrate that the results of limited sampling
were  inadequate to define the  volume  of "actionable"  con-
taminated soils. Kriging was also used to guide further sampling ef-
forts. In that evaluation, useful methods of presentation of the
statistical  results were developed. This method  of  presentation
makes the complex statistical results  more accessible  and effective
in  influencing the decision process.
   The case studies  support the  following  conclusions regarding
uncertainties associated with soils and subsurface contamination:
•Uncertainties regarding the adsorption process are severe and
  often allow no better  than order of magnitude estimates of con-
  taminant migration potential.
•Uncertainties regarding the permeability of soils can also lead to
  substantial  uncertainties in  estimating contaminant migration,
  though these uncertainties are typically less serious than those
  regarding adsorption.
 •When the effects of uncertainty in  both adsorption and perme-
  ability are combined,  extreme uncertainties in contaminant mi-
  gration estimates must be expected and recognized.
•The uncertainty in many environmental variables appears to be
 best described by a log-normal distribution; although these data
 were not presented in this paper, such distributions are apparent
 for contaminant concentrations,  alternative estimates of adsorp-
 tion coefficients and measurements of soils permeability in the
 same hydrogeologic unit. Log-transformation of environmental
 data frequently permits  the application of statistical procedures
 which have been derived upon the assumption of normality.

REFERENCES

I.  Freeze, R.A. and Cherry, /.A..  Groundwaler,  Prentice-Hall, Inc.,
   Englewood Cliffs, NJ,  1979.
2.  Brady, N.C., The Nature of Properties of Soils.  MacMillan Publish-
   ing Co., Inc., New York, NY, 1974.
3.  Lyman, W.J., Reehl, W.F. and Rosenblatt, D.H., Handbook of Chem-
   ical Property Estimation Methods: Environment*! Behavior of Organic
   Compounds. McGraw-Hill Book Co., New York,  NY. 1982.
4.  Karickhoff,  S.W.. Brown, D.S. and Scott, T.A., "Sorption of Hy-
   drophobic Pollutants on Natural Sediments," Water Research  J.. 13,
   1979, 249-255.
5.  Skrivan, J.A. and Kar linger.  M.R.,  Semi-Variogram Estimation and
   Universal  Kriging. USGS  Water Resources Division,  Tacoma, WA,
   1979.
6.  Gambolati, G. and Volpi, G.. "Groundwater Contour Mapping in
   Venice by  Stochastic Interpolators: 1. Theory," Water Resources Re-
   search. 155,  1979. 281-290.
312       RISK ASSESSMENT/DECISION ANALYSIS

-------
 DESIGNING RISK  ANALYSES TO AVOID PITFALLS IN  COST
    RECOVERY ACTIONS:  A LEGAL/TECHNICAL  SOLUTION
                                              JOHN C.  HALL, J.D.
                                         Wickwire, Gavin & Gibbs, P.C.
                                                 Washington, B.C.
 INTRODUCTION

  Over the past decade,  a significant amount of legislation has
 been passed to protect the American public from the adverse effects
 of toxic substances.1 Some of these laws are relatively easy to im-
 plement since Congress provided clear legislative tests to determine
 when and to what extent regulation is appropriate. For example,
 the Federal Food, Drug and Cosmetic Act avoids many difficult
 regulatory decisions by forbidding the intentional addition of any
 substance that has been shown to induce cancer in humans or other
 animals.2 The justification to regulate under this Act need not in-
 clude a determination that the carcinogenic substance poses an un-
 reasonable risk, because the statute establishes a no-risk standard.
 The applicability of animal bioassays to humans is not questioned,
 and the FDA does not have to analyze whether cancers  are likely to
 occur at doses below the  level of exposure that caused cancers in
 the animal tests. The basis for regulating under this statute is rather
 straightforward.
  CERCLA,3 which is the most recent legislation addressing toxic
 and hazardous wastes, adopts a similar direct approach. Through
 implementing guidance and judicial decisions, it has become evi-
 dent that responsible parties/defendants bear a considerable bur-
 den to avoid the reach  of the statute." Superfund authorizes gov-
 ernmental action  which substantially impacts regulated entities in
 the interest of protecting  the public health and permits institution
 of lawsuits  against the  culpable parties to recover the reasonable
 costs of solving the waste problem. It is significant that government
 is granted broad discretion to exercise its judgment5 in performing
 risk analyses to determine whether an imminent and substantial en-
 dangerment exists and in  developing solutions to these problems.
 These governmental determinations are  limited only  by the Na-
 tional Contingency Plan,6 which  becomes the basis for justifica-
 tion for expenditures. The responsible parties/defendants have lim-
 ited opportunity to challenge these costs and far less opportunity
 to attack the bases for such expenditures, i.e., the risk analyses
 performed by the government and its contractors.
  Although CERCLA provides a strong regulatory framework for
 addressing  threats posed by hazardous wastes, the "how  clean is
 clean" issue still presents a vexing problem to the USEPA. Un-
 like the water or air programs where a single medium is being
 assessed, Superfund sites typically have several routes of exposure;
 single media standards  are of little use. Attempts to address this
 issue have centered on use of site specific risk analyses that assess
 all routes of exposure to quantify the magnitude of the  threat pres-
 ent  and evaluate the reliability  and protection offered by various
response alternatives. This type of multi-media analysis is on the
cutting edge of science and, consequently, numerous questionable
assumptions are used to arrive at a final decision. It is not unusual
for only a few assumptions to make the difference between choos-
ing a no action alternative and a $10 million response action. How-
ever, for the USEPA to cover the cost of a response action based
on such assumptions, a  cost recovery  action is necessary. Al-
though there has not yet been a sizeable body of case law, the
courts are expected to accord substantial deference in review of dis-
putes risk analyses if the recent trend of Superfund case law con-
tinues. To gain insight into how courts may review Superfund risk
analyses in a cost recovery action, a review of court decisions in-
volving similar regulatory actions is appropriate.
  A  review of most other environmental statutes regulating toxi-
cant  levels in  the work place, home or environment reveals that
most statutes  are similar  to the  Superfund program; a threshold
risk should be present to justify initiating Federal response. These
statutes  usually require a  determination of unreasonable risk and
an evaluation  of the economic consequences of abating the "un-
reasonable" risk.7 Such statutes include the Consumer Product
Safety Act (CPSA),8 Resource Conservation and Recovery Act of
1976 (RCRA),' Occupational Safety  and Health Act of 1970
(OSHA)10 and the Toxic Substance Control Act (TSCA).''
  It is important to realize that  most toxics statutes are quite re-
cent developments. Prior to their enactment, agencies did not focus
resources to develop the expertise and  detailed information re-
quired to conduct comprehensive analyses of toxic substance im-
pacts. In the early case law, the courts realized that many issues in
the toxics field were "on the frontiers  of science", and,  there-
fore, gave substantial deference to agency decisions if there was an
apparent need to regulate.1J
  Substantial  deference was  not only characteristic in reviewing
risk analyses but alst) in conducting other complex environmental
assessments. One fairly exhaustive review of cases involving com-
plex  environmental models used in air, water and noise pollution
projection concludes that "courts in reviewing environmental cases
involving computer models and  other  quantitative methodologies
have been constrained to  apply a [sic] too limited and differential
standard of review in those decisions."13 This reflected the judi-
ciary's lack of expertise and exposure to the complicated problems
these analyses attempt to evaluate.
  However, a review of the recent case law indicates that courts are
now  willing to conduct a  more searching and thorough review of
the record upon which an agency has based  its decision.'4 This
trend is no doubt a sign that the  regulated community has become
more astute in dissecting agency analyses and perfecting its attacks
on complex reasoning. The judiciary has become either more famil-
iar with the topics or more sympathetic to the arguments raised by
                                                                    RISK ASSESSMENT/DECISION ANALYSIS
                                                      313

-------
those subject to regulation. We can expect nothing less in consider-
ation of the anticipated challenges to decisions  (and their costs)
prompted by risk analyses performed under Superfund.
  Since demands for accurate risk assessments have increased, sig-
nificant scientific improvements in  testing exposures and effects of
toxic substances are being developed and verified  at a  rapid pace.11
As  a result of the  recent expansion in scientific understanding,
some courts appear to believe that risk evaluation has become a
"precise" science and that demands upon the regulatory  agency for
detailed technical justification should, therefore, be increased."
However, this assessment of the state of the art is superficial if the
volume of law review articles on uncertainty in risk assessments is
an accurate barometer of scientific capabilities.''
  The  purposes of this paper are to provide an  indication of the
level of judicial review that one may expect for the various com-
ponents of a risk analysis performed for a Superfund site and to
put into context some important issues  that are  likely  to arise in
Superfund enforcement and remedial action cases. In this paper the
author first reviews the  general standard of judicial review appli-
cable to most risk analyses and then goes step by step through the
factors necessary to formulate a thorough risk analysis. The author
also evaluates the specific standard of review for each step. The
following pieces of the risk assessment puzzle will be explored:
•Data Base
•Model Selection
•Exposure Assessment
•Demonstration of Risk
•Quantification of Risk
  For each component of the risk analysis, a review of existing case
law will be presented and compared with the findings of a recent
judicial decision in the risk assessment field, Gulf South  Insulation
etal. v Consumer Product Safety Commission, (the "Foam Insula-
tion" Case)." This case  is used for  comparison because  it provides
an excellent example of  how thorough courts may be  in reviewing
risk analyses. To  briefly summarize,  the  Gulf South case  was
brought in response to a proposed ban on urea-formaldehyde insu-
lation. In April, 1982, after a six year investigation into  the effects
of  urea-formaldehyde foam insulation  (UFFI), the  Consumer
Product Safety Commission (CPSC) issued a  final rule  banning
UFFI in residences and schools.  The CPSC found that UFFI posed
an unreasonable risk of acute irritant effects and cancer. The Fifth
Circuit's review of this rulemaking  evaluated virtually every factor
of the CPSC risk analysis  for UFFI and concluded that  the analy-
ses and methodology were insufficient  to  support the necessary
finding that an "unreasonable risk" was present.

STANDARD OF REVIEW

  Prior to an investigation of an agency's decision-making record,
the court must determine the appropriate  standard of review to
apply.  Some statutes regulating carcinogens specifically state the
level of judicial scrutiny which agency regulations must pass.  For
example, section 2060 (C) of the CPSA states that a consumer pro-
duct safety rule shall not  be affirmed  "unless the Commission's
findings...are  supported by substantial evidence on the record
taken as a whole."  In Industrial Union Department. AFL-CIO v.
Hogdson, the court states  that under the substantial evidence test,
"our review basically must determine whether the Secretary carried
out his legislative task in a manner reasonable under the state of the
record  before him."" This is the concept of "reasoned decision-
making" which has evolved as the common standard for judicial
review of agency regulatory action."
  In determining whether or not a  decision  will meet the reasoned
decisionmaking test, the court assesses the agency's action in light
of the  statutory directive under which the regulation or action  is
promulgated." In certain situations, the statutory language or  leg-
islative history may be used by the agency as partial justification for
deciding an issue or interpreting data when two  or more answers
are plausible." For instance, a decision to regulate an uncertain
cancer  risk is more likely to be affirmed if the statute specifies that
"an ample margin of safety to protect public health"" is required,
                                                        than if the statute requires "that necessary to protect human health
                                                        and  the environment",  all  other  factors  being equal."  Latin
                                                        stresses that this is especially true when evaluating decision mak-
                                                        ing under uncertainty." In American Textile Manufacturers Insti-
                                                        tution v. Donovan (the Cotton Dust Case), the Supreme Court held
                                                        that  the Occupational Safety and Health Act defined the balance
                                                        that Congress desired when balancing risks to workers and costs."
                                                        Worker health was placed above all other considerations. Thus, the
                                                        court allowed the regulatory authority to err on the side of overpro-
                                                        tection when the information presented was conflicting or not con-
                                                        clusive.
                                                          The court in the Foam Insulation Case described this level of re-
                                                        view in  more detail.  "The facts that detract from the Agency as
                                                        well as those that  support  it are to be considered (cites omitted)...
                                                        the ultimate question is whether the record contains such relevant
                                                        evidence as a reasonable mind might accept as adequate to support
                                                        a conclusion (cite omitted)."" Review of the UFFI decision shows
                                                        that the Fifth Circuit applied the substantial evidence/reasoned de-
                                                        cision-making test to each and every decision.
                                                          Although one cannot voice disagreement with the Fifth Circuit's
                                                        choice of judicial standard of review, most courts apply this test to
                                                        the record  as a whole, and not to each and every point of the
                                                        CPSC analysis. To do so  amounts to a substitution of judgment
                                                        by the reviewing court; such action has been strongly discouraged
                                                        by the Supreme Court since Vermont Yankee." However, the Fust
                                                        Circuit  decision  appears  to  appreciate  that   numerous  small
                                                        "errors" in a risk analysis can rapidly compound and produce an
                                                        unreasonable result. For this reason, a more  thorough review may
                                                        be justified  if the record supports it.
                                                          In reviewing the application of the substantial evidence/reasoned
                                                        decision making test in the Cotton Dust Case, the Supreme Court
                                                        states that "the possibility of drawing two inconsistent conclusions
                                                        from  the evidence does not prevent the administrative agency's
                                                        finding from  being supported by substantial evidence (cite  omit-
                                                        ted)."" If the test were not applied in this fashion, it would require
                                                        almost flawless decisionmaking  by the regulatory agency. This re-
                                                        quirement,  of course, would be virtually impossible to meet.
                                                          In reviewing the OSHA analysis finding that a carcinogenic sub-
                                                        stance posed an unreasonable risk, the Supreme Court held in the
                                                        Benzene Case that the substantial evidence test requires an agency
                                                        to prove that the specific level of exposure "more likely than not"
                                                        presents "a significant risk of material health impairment."" In
                                                        this case, the Agency could not  demonstrate  that the existing stan-
                                                        dard posed  an unreasonable risk  of a material health impairment.
                                                          A recent  article on the  review of reasoned decision-making and
                                                        the substantial evidence test as applied to risk analysis suggests that
                                                        the standard "imposes three primary responsibilities on an agency
                                                        assessing health risks: (1) it must adequately evaluate the technical
                                                        data; (2) it must follow proper administrative procedures; and (3) it
                                                        must correctly carry out its statutory mandate."" This scheme re-
                                                        flects the framework that the Supreme Court established in Citizens
                                                        to Preserve Overton Park, Inc. v Volpe, a framework that was
                                                        structured to accommodate science-policy decisions."
                                                          Most federal courts of appeal  have declined to apply the substan-
                                                        tial evidence test to certain "legislative like" policy decisions that
                                                        must be made in a risk assessment, such as choice of an allowable
                                                        exposure level." The logic behind this decision is that when an
                                                        agency is faced with a lack of available evidence to specifically sup-
                                                        port one value over another, a "science-policy"  decision has to be
                                                        made. These  specific decisions are not amenable to review under
                                                        the substantial evidence test since they are not issues of fact. In this
                                                        situation,  a  rational  basis  test  similar  to the  "arbitrary and
                                                        capricious" standard of judicial  review is applied." This distinction
                                                        in the type of decision being made is critical in the hazardous waste
                                                        area  because  data on effects of substances  are often sparse and
                                                        final decisions are based on a combination of scientific fact, legis-
                                                        lative mandate and agency policy.
                                                           From the available case law, it appears that courts apply a flex-
                                                        ible standard of review depending upon the type of agency de-
                                                        cision. In deciding whether or  not an  agency decision meets the
314
RISK ASSESSMENT/DECISION ANALYSIS

-------
"substantial evidence/reasoned decision-making test," one must
first decide what type of issue is present (factual or legislative/pol-
icy) prior to applying the test. When legislative/policy decisions are
involved, greater deference is given. This approach to review of
agency decision-making recognizes that many regulatory decisions
are comprised of factual and non-factual decisions, and that the
standard of review should reflect the nature of the agency's de-
cision.

DATABASE

   A risk analysis, like any other scientific analysis, must be based
on reliable scientific information to  produce reliable results. The
agency should be convinced that the data are not biased, are appro-
priate to the situation being analyzed and are the best that can be
readily obtained. When data are conflicting, the agency should ex-
plain why the data were chosen  and other data omitted. Where the
agency has explained its rationale for selecting data, courts typical-
ly allow the regulatory agency significant discretion because of the
complexity of issues influencing the decision on appropriate data.35
   In most risk analyses, significant data gaps are present.  For this
reason, a major issue on review is often whether or not the data
base is sufficient to draw a reliable conclusion. In the Cotton Dust
 Case, the Supreme  Court refused to overturn OSHA's regulation
 on the claim  that various industry and private consultant cost esti-
 mates were all unreliable. The Court found that the precision of the
 estimates was limited by the absence of adequate industry data to
 develop more reliable estimates. Because the agency could not ob-
 tain more data, it acted reasonably in using what it believed to be
 the most reliable data." This position has been espoused on num-
 erous occasions by the lower Federal courts. Where the agency
 "recognizes" the thinness of the data base and explains "the rea-
 sons and its efforts to compensate therefore," the courts will not
 second-guess the agency.37 If deference was not given in this area,
 decisions to regulate toxicants would rarely survive judicial review
 due to lack of data.
   In the Foam Insulation Case, the court stated that there were
 numerous problems with the data base. These included: (1) the risk
 assessment was based on a biased sample of "complaint homes,"
 (2) the Commission erred in relying  exclusively on rat data for its
 risk assessment model and (3) the Commission ignored numerous
 epidemiological studies that indicate formaldehyde is not a human
 carcinogen.3' In addition, the court stated that the Commission did
 not explain its reliance on a data base comprised largely  of com-
 plaint homes.3' The court concluded the studies were inadequate to
 serve as a data base for the risk assessment.40
   If these concerns were accurate,  potentially there would be  a
 serious problem with the rule-making. However, the Commission
 did  explain its use of the data base very carefully, including its rea-
 sons for relying on the rat data (though it did not use it exclusively
 as the court  stated) and its rationale for finding the existing epi-
 demiological studies unpersuasive.4'  The apparently flawed review
 by the Fifth Circuit underscores the reason why most courts do not
 rigorously assess the agency's selection of the data base unless
 major flaws are apparent. Given the court's limited personnel, ex-
 pertise and time for review, one can hardly expect it to become as
 familiar as the regulatory agency with the strengths and weaknesses
 of the individual pieces of information presented to the agency for
 consideration.
   Even if there were substantial issues surrounding the choice of
 data used in an analysis, courts rarely substitute their judgment for
 the  views of  the agency on the relative merits of the data, so long
 as the agency has explained its reasons for using the data. In the
 Lead Case, the D.C. Circuit Court stated that "where the Agency
 presents scientifically respectable evidence which the petitioner can
 continually dispute with  rival...evidence, the court  must not
 second-guess  the particular way the Agency chooses to deal with
 the conflicting evidence and resolve the dispute."42 Other Circuits
 also hold similar views.43
   Thus, case analysis indicates that rather broad discretion is per-
mitted an agency in its selection of appropriate data; however, that
selection must be accompanied by a detailed explanation of how
and why these data were chosen. Actually, this amounts to no more
than good scientific documentation that would be expected of any
reliable technical analysis.

MODEL SELECTION

  In every quantitative environmental assessment,  including risk
assessment,  an  appropriate model must be chosen to characterize
the data and project results. Models range from rather simple, in-
corporating only a few major factors that influence the phenom-
enon, to extremely  complex, incorporating many factors. This
characterization should not lead the reader to believe that the more
complex model will always yield better results. Depending upon the
complexity of the situation and the decision to be made, the proper
choice of model will vary.
  The inherent assumptions found in the model will, at times, be
subject to judicial review. Assumptions may be built into the model
to decrease its complexity or may result from a lack of knowledge
concerning the  phenomenon being modeled. The distinction is im-
portant because greater deference will be granted to decisions based
on uncertainty than to decisions based on convenience.44
  Similar to the deference granted an agency in selection of a data
base, an agency's selection of the appropriate model is rarely ques-
tioned by the courts. In BASF Wyandotte Corp. v. Costle, the First
Circuit states, "we agree with the Fourth Circuit, the choice of sta-
tistical methods is a matter left to the sound discretion of the Ad-
ministrator. .. .the choice of any given method may mean that an al-
ternative would yield  different  results. The necessary corollary,
however, is that any other system chosen would be left to the same
criticism. We will not leave the Agency so vulnerable."45
  An important consideration in judicial review of model selection
is the degree of scientific knowledge available to model the subject
in question. Lack of knowledge in the risk assessment area may re-
quire the agency to make numerous scientific assumptions, not all
supportable  with  substantial  evidence. As stated in Industrial
Union Dept. v. Hodgdon,  "where existing method or research is
deficient, the agency enjoys broad discretion (to  regulate) on the
best available information."46 When there exists uncertainty over
the type of model to use, the decision to choose one model over an-
other is a  science-policy decision,  not a factual one. As  such,
greater deference should be accorded the decision.
  The court in the Foam Insulation Case goes substantially beyond
the level of review that other Circuit courts invoke when reviewing
the choice of model. The court stated that use of a "no threshold"
model to predict the likely incidence of cancer is "of questionable
validity."47  This conclusion is reached in spite of the fact that the
National Academy of  Sciences Committee on  Toxicology con-
cluded that there was no threshold for the irritant effects of formal-
dehyde.48 In addition, the Chemical Industry Institute for Toxicol-
ogy study,  reviewed  by the Federal Panel on Formaldehyde, con-
cluded that  formaldehyde should be presumed to pose a carcino-
genic risk to humans.4' It is also commonly accepted that carcin-
ogens do not have a threshold exposure level since science is not
capable of adequately defining  it  at this  time.50 Based on these
facts, one would hardly describe  CPSC's decision to use a no thres-
hold model as  "questionable."  Rather, it appears to be well rea-
soned and supported by a body of respectable science. The Fifth
Circuit's comments on the model selection do not appear to be well
founded.
  One may conclude from the cases reviewed that the choice of a
specific risk assessment methodology will be rarely second-guessed
as long as the applicability of the model is reviewed, the primary
variables are accounted for and  the ability to make better projec-
tions is discussed.51

EXPOSURE ASSESSMENT

  An exposure assessment is the phase of a risk  analysis that at-
tempts to evaluate the various routes and conditions under  which
                                                                           RISK ASSESSMENT/DECISION ANALYSIS
                                                          315

-------
exposure occurs or is likely to occur. The level of total exposure to
the toxicant is estimated at this point. Once the total exposure  is
calculated, the effect from that exposure may be quantified by
comparing the result to the risk assessment model output. Since an
exposure assessment is primarily a factual determination, substan-
tial evidence must support the findings.
  As the following case  analysis will demonstrate, courts are more
comfortable questioning this part of a risk analysis than any other.
This appears to occur because exposure analyses are easy for lay-
men to understand and are not particularly  technical in nature.
From our own experience, we can assess whether or not a particu-
lar exposure scenario is  reasonable.  For example, assuming that a
person stands at the maximum point of concentration 24 hr a day,
365 days a year for 70  yr sounds highly improbable. One would
question, what about rainy days, or sending the children to school
seven hours a day for 12 yr? Should not these  factors require an
adjustment to the predicted exposure? Due to  a common sense ap-
proach  that is applicable to this part of the  risk analysis, courts
are not hesitant to question the exposure analysis.
  The first step in the exposure assessment is evaluating the routes
of exposure. The primary routes of human exposure include inges-
tion, absorption and inhalation. In the Benzene Case, the Supreme
Court rejected the dermal contact standard set by OSHA that was
intended to reduce  potential cancers caused by absorption." The
Court found that OSHA's exposure assessment for this route was
"based on the assumption that [benzene] can be absorbed through
the skin in sufficient amounts to present a carcinogenic risk" rather
than on evidence showing that absorption may actually occur."
The lower court found that "the record fails  to demonstrate that
benzene is absorbed through the skin."94 Thus, it appears that the
regulatory authority must establish with substantial evidence that
an exposure route is reasonably likely to occur before including it in
the exposure assessment.
  Another concern in reviewing the exposure assessment is assur-
ing that the population to be protected has a level of exposure sim-
ilar to the population measured. In Texas Independent Cinners
Association v Marshall,  OSHA tried to apply a risk analysis devel-
oped for different workers at different exposure levels  where  the
only similarity was the route of exposure that  was present." How-
ever, the Fifth Circuit refused to apply an OSHA  standard devel-
oped from data and analyses based on cotton dust levels inside mills
to dust levels in ginning  operations. Due to "significantly different
conditions and significantly different exposure levels" for the gin-
ning workers, the court held that OSHA failed to demonstrate that
existing working conditions posed a serious health threat."
   The factors evaluated in the Texas Ginnery case were very sim-
ilar to the factors considered in the "technology transfer" cases
under the Clean Water  and Clean Air  Acts." In  these cases, the
courts required the  USEPA to prove that technologies used in one
type of operation will also be feasible and will achieve the proposed
effluent reductions in another industrial  application. Normally,
significant similarities such as composition  of waste,  strength,
process variability, etc., must be demonstrated between the differ-
ent industrial wastes and operations before the court  will accept
the agency's conclusion  that the analysis conducted for one group
is applicable to another." There is, however, considerable prece-
dent where an agency data transfer has been upheld, even when
the applications are dissimilar. In Reserve  Mining  Company  v
EPA, the court permitted  the  USEPA to regulate discharges of
asbestos fibers into the  waters of Lake Superior based on inhaled
asbestos levels known to cause serious harm."  Clearly, this was not
the same route of exposure known to  cause  cancers nor was the
level of exposure reliably calculated to be the same proven to cause
cancers." The action was supported as "a precautionary and pre-
ventive measure to protect public health."61
   Overestimating the exposure level may also invalidate the agen-
cy's risk analysis. In the Foam Insulation Case, the court stated
that "the failure to quantify the risk at  the exposure actually asso-
ciated with UFFI is the finding's Achilles heel."" The court in-
dicates  that CPSC had  overcalculated the potential exposure that
                                                        would normally be encountered because they relied primarily on
                                                        data from complaint houses." The court also stated that the con-
                                                        ditions tested "reflected conditions similar to an unheated, unair-
                                                        conditioned home, not an average home."" Finally, the court re-
                                                        jected the extension of the UFFI ban to schools because pergont at
                                                        academic institutions would not  have the  same duration of ex-
                                                        posure assumed in the risk assessment model (16 hr a day, 7 days a
                                                        week for 9 yr)." Thus, the court  points out several important re-
                                                        quirements for the exposure assessment: (1) accurate reproduction
                                                        of exposure level for subject population, (2) similar conditions to
                                                        "real world situations" and (3),  in the case of a "transferred"
                                                        analysis, relatively similar conditions (i.e., level and duration of ex-
                                                        posure).
                                                          Although the court in the UFFI case does point out valid areas of
                                                        concern, its factual analysis missed the mark. As previously stated,
                                                        CPSC did  not,  in  fact,  rely  primarily on  complaint homes to
                                                        develop the exposure level. The Commission undertook a rigorous
                                                        statistical analysis to determine whether or not complaint homes
                                                        had significantly higher levels of formaldehyde than non-complaint
                                                        homes. The analysis indicated that the complaint homes did not
                                                        have a significantly higher level. Thus, use of complaint home data
                                                        would not  significantly overestimate the calculated  in-home ex-
                                                        posure levels."
                                                          In addition, the tests run in the Franklin/Oak Ridge laboratories
                                                        were not similar  to "an unheated, unairconditioned home"; they
                                                        were run indoors in heat controlled settings.'7 One must also find it
                                                        unusual that the court, as a matter of law, concluded that the aver-
                                                        age house has air-conditioning and that the air-conditioning is used
                                                        regularly.
                                                          In support of the decision to vacate the ban on installing UFFI
                                                        in  schools,  one could reasonably rely on the Texas Ginners be-
                                                        cause CPSC did  not have data demonstrating that  use in schools
                                                        was hazardous. However, under the Reserve Mining approach, the
                                                        ban does not seem unreasonable since a sensitive segment of the
                                                        general public would be at risk (i.e., children) and UFFI is known
                                                        to cause acute irritant effects to approximately 20% of the healthy
                                                        individuals  exposed to levels typically  encountered  in the first 30
                                                        weeks after installation.*' In  addition,  children living  in UFFI
                                                        homes and also attending UFFI schools would receive significantly
                                                        higher doses than projected by the risk assessment. Thus,  their risk
                                                        would be substantially underestimated rather than overestimated as
                                                        in the Texas Ginnercase."
                                                          Generally, it appears that exposure assessments must be sup-
                                                        ported by substantial evidence to pass judicial review. However,
                                                        there appears to  be an exception when  reviewing uncertain ex-
                                                        posures. If  the public health is involved and a significant number
                                                        of persons are potentially affected, courts are more likely to err on
                                                        the side of safety and allow regulation. Where small discrete pop-
                                                        ulations  and significant expense  are  involved in  regulating the
                                                        potential exposure, courts tend  to require more specific verified ex-
                                                        posure analyses to justify regulatory action.  The exposure analysis
                                                        review in the Foam Insulation Case highlights an area which is like-
                                                        ly to receive more attention in  future cases; the reasonableness of
                                                        the exposure scenario. Compounding conservative assumptions can
                                                        easily result in the analysis overpredicting the likely exposure by a
                                                        factor of a million. This does not imply that a "worst case" analy-
                                                        sis is not acceptable; however, it should be a reasonable worst case.

                                                        DEMONSTRATING A RISK
                                                          After an exposure assessment is completed, the  next step is to
                                                        evaluate whether or  not a significant  risk is  present. To demon-
                                                        strate that a risk exists, the agency must have sufficient evidence to
                                                        demonstrate that current or anticipated exposure levels pose an un-
                                                        reasonable  risk.  Such data may include health studies, epidem-
                                                        iological studies or animal studies. The demonstrated risk may be
                                                        acute health effects, cancers or subchronic disorders to name a few.
                                                          Normally, the agency will be presented with numerous conflict-
                                                        ing reports on the level and likelihood of harm caused by exposure
                                                        to the toxicant(s) in question. The agency must determine from this
316
RISK ASSESSMENT/DECISION ANALYSIS

-------
information whether or not the harm is significant and likely to
occur given the circumstances of the exposure assessment. Courts
tend to be rather deferential in this determination. The Supreme
Court has stated that when available evidence of equivalent quality
is conflicting, a finding in accordance with one view or the other
should be considered to be supported by substantial evidence.70
  Recent cases show a trend toward requiring an agency to demon-
strate both qualitatively and quantitatively that a risk exists;  a
risk assessment analysis is almost a necessity for any agency wish-
ing to defend new  toxics regulations.  In the Benzene Case, the
Supreme Court stated that OSHA must prove a specific level of ex-
posure "more likely than not" presents "a significant risk of ma-
terial health impairment."" In the Benzene Case, OSHA did not
present data that the current 10 ppm standard posed any health
hazard.  OSHA relied primarily on epidemiological data taken at
much higher levels (estimated to be about 100  ppm) and the prem-
ise that no level of exposure was safe.72 The Court rejected this as
rulemaking  based  on assumption  and  stated  that  regulations
"based  only on inferences drawn from studies involving much
higher exposure levels rather than on studies involving these levels
or sound statistical projects from the high level  studies "...are not
adequate to demonstrate that a significant risk exists.73 The Court
also stated  that the  agency could base its determination of signif-
icant risk on animal studies conducted at higher  levels if a dose re-
sponse curve was generated.74 Other courts have also held that ex-
trapolation from animal data may be used to establish the prob-
ability of harm to man.75
   When faced with uncertainty over the effects of a carcinogen at a
specific level, agency action is not precluded if substantial evidence
exists to demonstrate a risk. In the Lead Case,  the court stated that
"the Agency decision may be fully supportable if it  is based on the
inconclusive but suggestive results of  numerous studies."76 The
court also went on to evaluate the Supreme Court's decision in the
Benzene Case. It indicated that in the plurality's view, the OSHA
standard "rested on rigid categorical assumptions about the health
dangers  of  benzene,  rather than  specific evidence of its likely
harm."77 The  Benzene Case  also  supports  the D.C. Circuits'
position  that a precise correlation of exposure and effect is not
necessary to revise an existing standard. However, the agency must
present evidence "showing it is more likely than not that there is a
significant risk" at the existing standard.78
   The standard set  in  the Lead Case, unlike  most other toxicant
standards reviewed by the courts, was based on subclinical effects
rather than the likelihood of cancer.79 Protection from  subclinical
effects was also the basis for  the standard set by  OSHA in the
Cotton Dust Case'0 that was sustained after Supreme Court re-
view." These cases  demonstrate that  non-cancer  health impair-
ments may also be considered significant risks.
   A review of the scientific justification required by the Fifth Cir-
cuit to demonstrate  that a significant cancer risk was present, indi-
cates that the Fifth  Circuit's decision is substantially at odds with
prior Supreme Court decisions  and those expressed by other Cir-
cuits. The court implied that CPSC's evidence  of  carcinogenicity
was insufficient to demonstrate significant risk  because, although
animal studies produced nasal cancers,  the human epidemiologic
studies did  not find  a statistical increase in such  cancers.82 In  addi-
tion, the court stated that a single animal study based on 1920 sub-
jects (the court erroneously stated only 240 subjects were used)  is
insufficient to develop a dose response curve  for formaldehyde.83
These findings  are contrary to scientific standards established by
leading government cancer authorities  and ignore  the substantial
body of evidence indicating that formaldehyde is carcinogenic.'4
  This outright substitution of judgment is also contrary to the
Fifth Circuit's  own language on the scope of  judicial review of
agency findings enunciated in  American Petroleum Institute v.
OSHA."
  As a parting blow, the court stated that using animal data to sup-
port a  finding of unreasonable cancer risk  is  "of questionable
validity" since the industry points out that "...it is far more sen-
sible to assume that rats equal mice than that rats equal humans. "86
If the court means to imply that human test data at current con-
centrations are necessary to substantiate a present health risk, one
would have a difficult time finding support for this position in any
Supreme Court or other circuit court decision.87


QUANTIFICATION OF RISK

  To determine whether a demonstrated risk is significant, some
quantification of risk is normally required. The Supreme Court's
decision in the Benzene Case has generally been interpreted to im-
ply that risks must be quantified and that incremental benefits of
reducing exposure must be assessed.88 This belief appears to result
from the Court's directive that an agency must prove that a specific
level of exposure "more likely  than not" presents "a significant
risk of material health impairment."89 However, other language in
the Benzene Case indicates that quantitative estimates are  not re-
quired, especially "when at the frontiers of science.'"0 Thus, the
Court recognizes that situations will exist where  a  numerical  risk
analysis cannot be conducted with any substantial  degree  of reli-
ability. In addition, the Court states, "although the agency has no
duty to calculate the exact probability of harm, it does have an ob-
ligation to find that a significant risk is present before it can char-
acterize a place of employment as unsafe."" Recent cases show,
however, that quantifying the risk at the existing levels of exposure
will increase the agency's chances of successfully defending its regu-
latory actions.
  In the Cotton Dust Case, the risks  of bysinossis from cotton dust
were well established, and OSHA was able to construct a dose/re-
sponse curve from epidemiological data. The agency analyzed the
harm at  current levels and also at  the proposed permissible ex-
posure limit (PEL). The  rulemaking, as related to  the new PEL,
was upheld on all respects.92  The Lead Case also demonstrates the
D.C. Circuit's tendency to approve agency action when a quantifi-
cation of the risk from exposure and improvement from reduced
exposure has been conducted.  The  court noted  that, unlike  the
Benzene  Case, "OSHA was able to describe the actual harmful
affects of lead on worker populations at both the current PEL and
the new PEL" by using an air/blood level correlation.93 It appears
that an agency should  quantify  the risks and improvements asso-
ciated with the rulemaking whenever possible.
  When requiring a quantified  risk, most courts  appear to be re-
ferring to the existing condition as opposed to justifying the spe-
cific standard chosen below the unreasonable risk level. In the Ben-
zene Case, the Supreme Court's major complaint with the  OSHA
analysis was that it assumed  carcinogens should be reduced to the
lowest extent feasible.94 OSHA did not have any data demonstrat-
ing the presence of a risk at the current permissible exposure level
(PEL). Without data verifying that the current PEL posed a risk,
the agency should have developed a risk assessment model to trans-
late the effects of high exposure levels to the existing low exposure
levels. The Court points to numerous other OSHA actions where
OSHA quantified the number of lives saved or cancers prevented
as better examples of justified agency rulemaking.95
  In the  Coke Ovens Case, which followed the Benzene Case, the
court did not require an estimate of cancer risks at levels below
the old PEL.96 In  a  rulemaking involving vinyl chloride,  OSHA
had animal test data at the current PEL of 50 ppm showing that
50% of the animals died after 11 months. This was considered evi-
dence of a significant threat  at the current exposure level, and the
agency did not weigh the relative benefits of setting a new PEL
other than 1 ppm. The primary consideration at this point  was
whether or not the proposed PEL was "feasible."97
  Most courts have also taken the position that the final standard
chosen by a regulatory body will not undergo the same rigorous re-
view as the  methodology to arrive at that standard. "Where the
standard requires OSHA to set a numerical limit for some phenom-
enon we must remember that the precise choice of a number  is
essentially a 'legislative' judgment to which we must accord great
deference and  which  only must  fall  within  the  'zone  of
reasonableness', "98
                                                                          RISK ASSESSMENT/DECISION ANALYSIS
                                                          317

-------
  These cases reflect the deference the Supreme Court typically
accords agency decisions that attempt to quantify risks. "So long
as [findings] are supported  by a  body  of reputable  scientific
thought, the Agency is free to choose conservative assumptions in
interpreting data with respect to carcinogens, risking error on the
side of overprotection rather than underprotection.""
  Thus, it is apparent from the case law that regulation based on a
quantified risk analysis has a greater chance of passing judicial re-
view than a qualitative risk analysis;  however, quantitative analyses
are not strictly required to pass judicial muster. In addition, the
quantified risk analysis must demonstrate with substantial evidence
that the existing standard or condition poses an unreasonable risk
rather than that the proposed standard is the correct one.
  In the Foam Insulation Case,  the court took the general position
on risk quantification that other courts have taken.  They stated,
"predicting how  likely an  injury is to occur, at least in general
terms, is essential to a determination of whether a risk of injury is
unreasonable."100 The court also stated that it is necessary to quan-
tify risks at the exposure level  actually associated with the exist-
ing exposure. Beyond this point, the Fifth  Circuit significantly de-
parted from the other courts.
   In the Foam Insulation Case, CPSC appeared to follow the
directives laid down in the Benzene Case on quantifying risks; how-
ever, the court rejected the  analysis. The CPSC used a risk analysis
model to project the cancer risk posed by  UFFI in houses. Unlike
the Benzene Case, they had sufficient animal data at  several doses
and were able to construct a  dose/response  curve from that data.
The CPSC risk analysis predicted that the cancer risk was 0 to 51
per 1,000,000.
   Besides the data base issues previously discussed, the court stated
that the estimates were insufficient because they were not precise.'"
However, the Supreme Court has specifically stated that "precise"
estimates are not required.'02 The Fifth Circuit also indicated that
the "margin of error is inherently large," and thus the analysis re-
sults were  unreliable."" If the  court means that  uncertainty pre-
vents CPSC from acting, this too has been specifically rejected by
the Supreme Court.104 The  Fifth Circuit also implied that using the
"upper level of risk" does  not constitute substantial evidence that
"it is more likely than not the UFFI presents a significant risk of
cancer."105 This  line of reasoning  has also been rejected by the
Supreme Court since it has stated that the  agencies may err  on the
side of overprotection.
   Thus, based on a review of the existing case law, the Fifth Circuit
has gone significantly beyond the level of detail and justification
normally required to demonstrate  substantial carcinogenic risks.
Whether other jurisdictions are likely to follow this  case remains
uncertain.
HOW COURTS ARE LIKELY TO VIEW  RISK ANALYSES
PERFORMED UNDER SUPERFUND
   The foregoing analysis of judicial decisions in the hazardous
waste area indicates that substantial  deference is normally accorded
agency decisions  to regulate those substances. And,  as previously
noted, the enabling statute plays a central  role in determining how
agency decisions should be reviewed. Considering the broad grant
of powers established under CERCLA, it is  difficult  to argue that
Congress  intended  CERCLA actions to  be reviewed  any  more
rigorously than other agency decisions in the hazardous waste field.
However, unlike the other statutes relating to control of toxic sub-
stances, responsible parties under CERCLA are directly  liable for
costs incurred in abating a hazardous waste threat.  For this rea-
son, greater caution  should be exercised by the courts before find-
ing that a party is responsible and that all of the cleanup costs were
necessary to abate the hazardous waste threat.
  In most  civil cases involving  recovery of money damages, the
burden of proof that a plaintiff must carry is a "preponderance of
the evidence" (i.e., more likely than not). Although this burden of
proof is essentially  the same as the  "substantial  evidence/rea-
soned  decision-making test" as  applied to  factual risk analysis de-
cisions, in  areas of uncertainty, judicial deference allowed under
                                                        this test clearly does not equate with the burdens of proof under
                                                        the  "preponderance of  evidence"  test.  Thus,  it  is uncertain
                                                        whether courts are likely to review Superfund risk analyses more
                                                        closely. One could speculate that courts will still place a heavy bur-
                                                        den on defendants because of the possible threat to public health
                                                        that they  have created  and  because Congress,  by  granting the
                                                        USEPA board discretion under CERCLA, intended the courts to
                                                        be deferential to USEPA decisions to respond to hazardous sub-
                                                        stance releases. In  short, doubts concerning  impacts on  public
                                                        health are likely to be resolved on  the  side of protecting  public
                                                        health.

                                                        CONCLUSIONS
                                                          Risk analyses, like other environmental effects analyses, are re-
                                                        viewed primarily under the substantial evidence/reasoned decision-
                                                        making standard on the record. This has been interpreted to mean
                                                        that the regulatory  authority  must demonstrate that the existing
                                                        exposure it seeks to reduce  more likely than not presents a signif-
                                                        icant health risk.
                                                          In reviewing the analyses used to demonstrate the human health
                                                        or environmental risks, courts appear to be conducting more rigor-
                                                        ous  reviews than ever before. These reviews are tempered, how-
                                                        ever, by the statutory mandate under which the agency is acting and
                                                        a respect for the limits of scientific capabilities. These two factors
                                                        help the court to differentiate between factual decisions as opposed
                                                        to legislative/policy  decisions in the record. The latter are accorded
                                                        greater deference (almost an  arbitrary and capricious standard),
                                                        and the  former are  reviewed  more rigorously. Based on the case
                                                        law reviewed a risk analysis will be capable of passing the substan-
                                                        tial evidence/reasoned decision-making standard of review if the
                                                        following steps are followed.

                                                        Date Base
                                                        •Courts recognize  that conflicting information will  often be en-
                                                         countered; thus, be certain to present an adequate analysis of why
                                                         one set of data was chosen over another.
                                                        •Include a discussion of the quality of the data base and the ability
                                                         to  get "better" data (e.g., little help from industry being regu-
                                                         lated).
                                                        •When a particular subset of data is used that could be considered
                                                         biased or not representative, explain this choice.
                                                        •Indicate whether the data are likely to  produce overestimates or
                                                         underestimates.
                                                        •Where  the data are admittedly "thin",  explain the "policy" rea-
                                                         son or scientific reason for using the data as a basis for regulation.
                                                        Model Selection

                                                        •Where no specific model best fits the data, explain why one model
                                                         is chosen over another.  If it is a policy reason (i.e., a conserva-
                                                         tive model to insure protection), state so explicitly.
                                                        •Review the model assumptions  and determine whether they are
                                                         reasonable for the current application.
                                                        •If certain factors are not accounted for  in the model, explain why
                                                         they  are either not considered important or cannot be properly
                                                         characterized at this time.
                                                        •State what assumptions in the model are not supported by the data
                                                         and why the model is used in any event.
                                                        Exposure Assessment

                                                        •When choosing the routes of exposure,  demonstrate the actual or
                                                         likely existence of the route.
                                                        •If a previously undocumented exposure pathway is included or
                                                         serves as the primary bases for a new analysis, explain the basis for
                                                         the decision. This may be a legislative/policy rational, where sub-
                                                         stantial risk may be present and a large  population potentially ex-
                                                         posed but data do not substantiate that the pathway exists.
                                                        •Courts tend to view an exposure assessment as site specific. Only
                                                         use an analysis from another situation where a new assessment  is
                                                         not feasible.
318
RISK ASSESSMENT/DECISION ANALYSIS

-------
•When "transferring"  an exposure analysis to another group,
 demonstrate significant similarities for level and duration of ex-
 posure, route of exposure and characteristic of substance.
•Indicate the confidence in the exposure estimate, if possible (i.e.,
 is it likely to be an overestimate or underestimate?).  Also dis-
 cuss why more accurate assessments are not practicable.
•Where laboratory tests are used to simulate the rate of exposure,
 discuss differences/similarities between lab and actual conditions.
•Discuss how  widespread  exposure may be  and whether  it is  a
 threat to the general public.  This weighs in favor of taking action
 where precise exposures are less certain.
Demonstrating A Risk

•Present data or analyses that indicate current exposure levels pose
 an unreasonable risk.
•Where non-life threatening risks are protected, explain why.
•Animal tests should either be run at the current  exposure level or
 be suitable for use in a risk assessment model.
•Indicate how the data are similar to the exposure situation.
•If the type of illness found in the animal data is not similar to any
 known human illness caused by the carcinogen,  explain why this
 does not disqualify the data.
•Conduct an incremental quantitative evaluation of risk where pos-
 sible.
•If above a certain number of cancers per  1,000,000 is considered
 an unreasonable risk, explain why. This may be a policy rationale.
•Where epidemiological studies are used to  demonstrate risk, com-
 ment on their similarity to the present situation.
•Where a good correlation of exposure and effect cannot be made,
 explain why and whether or not more testing  is desirable (i.e., why
 the agency believes enough evidence is present).
Quantification of Risk

•Demonstration of significant risk will usually require a quantifica-
 tion of risk.
•Although quantifying  benefits obtained by reduced carcinogenic
 levels is not usually  required, it is a good idea to include this analy-
 sis if possible.
•If present risks cannot be quantified, provide a  detailed explana-
 tion of why, what  would be needed to reliably  conduct such an
 analysis and whether it is reasonable to wait for such an analysis.

Other Factors
•Wherever possible, indicate whether legislative/policy  decisions
 or  factual decisions are being made.  Factual decisions  generally
 require more supporting evidence.
•Where uncertainty  exists, rely on legislative mandate in addition
 to other inferences that may be drawn from the data.
•Small incremental changes in existing carcinogenic standards are
 usually supportable only if a lower standard would have been im-
 plemented but technology prevented implementation of that limit.
•Avoid the Fifth Circuit if you are a regulatory  body; this circuit
 clearly applies more stringent requirements than  any other. If you
 are the consumer advocate, race to the B.C. Circuit.
 REFERENCES

   1. See Gough, Laws for the Regulation of Carcinogens; Identifying and
     Estimating the Risks that  the Laws Seek to Reduce,  Toxic Sub-
     stances J. 4, 1983, 251.
   2. 21 U.S.C.  §348(c)(3)(A) (1976).  More  commonly known as the
     Delaney Clause.
   3. 42 U.S.C. 9601 (1980).
   4. In U.S. V. Chem-Dyne 572 F. Supp. 802,810 (1983), the court held
     that joint and severable liability applies under CERCLA unless the
     defendant demonstrates  a  reasonable  basis for apportionment;
     U.S. v. North Eastern Pharmaceutical  & Chemical Co.,  No. 80-
     5066 CVSW-4  (W.D. Mo.) (1984), holding both the president and
     vice-president of NEPACCO, the sole generator, personally liable;
    U.S. v.A& F Materials Co., No. 83-3123 (S.D. 111.), (1984) holding
    that CERCLA could be applied to past non-negligent off-site gener-
    ators.
 5.  Section 104(a)(l) of the Act authorizes the President to respond, con-
    sistent with the National Contingency Plan (NCP), whenever "any
    hazardous substance is released or there is a substantial threat of re-
    lease into the environment." Section 105  of CERCLA states that the
    NCP shall include "methods and criteria for determining the appro-
    priate extent of removal, remedy, and other measures authorized by
    this act." A review of the NCP indicates that many factors are con-
    sidered in determining the appropriate extent of remedy; however, a
    "significant" threat must usually be present prior to initiating a  fed-
    eral response action. See NCP §300.68 and §300.65. Section 106 of
    CERCLA also allows the President to seek relief where an imminent
    and substantial endangerment  is present. Such relief may take the
    form  of administrative orders or injunctions where appropriate.
    Demonstrating a significant threat typically requires some form of
    risk analysis.
 6.  Section  107 of CERCLA requires that the cost removal or remedial
    action incurred by a Federal or State government be "not inconsis-
    tent with the national contingency plan" to be recovered from a re-
    sponsible party. Considering the general nature of the NCP, it is not
    likely that many USEPA response actions will  be found inconsistent
    with the NCP.
 7.  See Note 1 at 254-255.
 8.  15 U.S.C. §§2051-2081(1976).
 9.  42 U.S.C. §§6901-6987(1978).
10.  29 U.S.C. §§651-678(1976).
11.  15 U.S.C. §§2601-2629 (1976 & Supp. v 1981).
12.  See generally Reserve Mining Co. et al.  v.  U.S., 514 F.2d 492 (8th
    Cir. 1975) (en  bane); Merrill, CPSC Regulation of Cancer Risks in
    Consumer Products:  1972-81,  67 VA. L. Rev. 1261 (1981); Case,
    Problems  in Judicial Review Arising From the Use of Computer
    Models and Other Quantitative Methodologies in Environmental De-
    cisionmaking,  10 Boston Col. Envtl. Affairs L.R. 251 (1982); Sub-
    stantive and  Procedural  Discretion  in  Administrative Resolution
    Science  Policy Questions: Regulating Carcinogens in  EPA  and
    OSHA, 67 Georgetown L. J. 729 (1979).
13.  Id, Case at 363.
14.  Compare,  Industrial Union Dep!, AFL-CIO v. American Petroleum
    Institute, 448  U.S. 607 (1980) (the Benzene Case) with  Industrial
    Union Dept, AFL-CIO v. Hodgson, 449 F.2d 467 (D.C. Cir. 1974)
    (the Asbestos Case). Also see generally Amer. Textile Mfr. 's Instil.
    v. Donovan, 452 U.S. 490 (1981) (the Cotton Dust Case) and United
    Steelworkers of America, Etc. v. Marshall, 647 F. 2d 1189 (D.C.
    Cir. 1980), cert, denied, 453 U.S. 913 (1981) (the Lead Case).
15.  See Wallace, Measuring Direct Individual Exposure to Toxic Sub-
    stances, Toxic Substances J. 4, 1983, 174.
16.  In the 5th Circuit Review of the Benzene Case, American Petroleum
    Instil, v. OSHA, 581 F.2d at 507 (1978), the court indicates  that
    dermal adsorbtion of benzene could have  been verified by using a
    radioactive tracer test to check the rate of benzene adsorption and for
    this reason OSHA did not use the "best data available."
17.  Latin, The "Significance" of Toxic Health Risks: An Essay on Legal
    Decisionmaking Under Uncertainty,  10 Ecology L.Q.  339 (1982);
    Rodgers, Judicial Review of Risk Assessments: The Role of Decision
    Theory  in  Unscrambling  the Benzene Decision,  11  Envt'l L.  301
    (1981);  Ashford et al., A Hard Look at Federal Regulation of
    Formaldehyde: A  Departure  From  Reasoned Decisionmaking, 7
    Harv. Envt'l L.R. 297 (1983).
18.  701 F.2d 1137 (5th Cir. 1983).
19.  449 F. 2d 467,474 (D.C. Cir. 1974).
20.  See NRDC v  EPA, 655 F.2d at 328  (1981)  and Note 14 Ashford
    etal.
21.  In American Paper Instil, v. EPA,  540 F.2d at 1028 (10th Cir. 1976)
    the court stated that all issues should be viewed in light of the Con-
    gressional intent.
22.  In E.D.F. v Costle, 578 F.2d 342-344 (D.C. Cir.  1978) the court re-
    lied heavily on the statutory language and legislative history to de-
    termine whether or not the EPA action was reasonable.
23.  Clean Air Act, 42 U.S.C. §7412(b)(l)(B).
                                                                             RISK ASSESSMENT/DECISION ANALYSIS       319

-------
24. Resource Conservation and Recovery Act, 42 U.S.C. §§6922,6924.
25. SeeNotel7,Iflrmat395.
26. 452 U.S. at494(1981).
27. 701F.2dat 1141 (1983).
28.  Vermont Yankee Corp. v. NRDC, 435 U.S. 519 (1978).
29. 452 U.S. 490, 523(1981).
30. Industrial Union Dept..  AFL-CIO v American Petroleum Instil., 448
    U.S. at 653, n. 63(1980).
31. See Note 17, Ashfordet al. at 305.
32. 401 U.S. 402 (1971).
33. See Note 17 Rodgers at 305 and O'Malley, American Petroleum In-
    stitute v Occupational Health and Safety Administration,  581  F.2d
    493 (5th Cir. 1978), lOEnv't'l L. at 673 (1980).
34. See Note 17, Ashfordet al. at 365, n. 462.
35. American Meat Institute v. EPA, 526 F.2d at 457 (7th Cir. 1975).
36. 452 U.S. at 528 (1981).
 37.  Weyerhauser Co.  v.  Costle, 590 F.  2d  1011, 1054 (D.C. Cir. 1978).
     Also see BSAF Wyandotte  Corp.  v. Costle, 598 F.2d 637, 653 (1st
     Cir. 1979).
 38.  701 F.2d at 1143(1983).
 39.  Id. at 1145.
40.  Id.
 41.  See Note 17, Ashford et al. at 363-368 for a review of these issues and
     the CPSC Petition for Rehearing filed May 5, 1983.
 42.  United Steelworkers of America v. Marshall, 647 F.2d at 1263 (1980).
 43.  See cases cited in Notes 35 and 37.
 44.  As stated earlier, courts tend to separate agency decisions into those
     that are factual and those that are legislative/policy in nature. Where
     a science/policy decision is made due to insufficient data, courts do
     not tend to question the expert agency decision.
 45.  598 F.2d at 655  (1979). The work of Mr. Case (See Footnote 12)
     also corroborates that courts give great deference in  reviewing de-
     cisions on selection of mathematical models or statistical techniques.
 46.  499 F.2d 474, n. 18(1974).
 47.  701F.2datll47, n. 20 (1983).
 48.  Id. at 1141.
 49.  Id. at 1146.
 50.  American Iron and Steel Inst. v. OSHA, 577 F.2d 825, 832 (3rd Cir.
     1978).
 51.  National Lime Association  v EPA, 627 F.2d at 452-53 (D.C. Cir.
     1980).
 52.  448 U.S. at 661 (1980).
 53.  Mat662.
 54.  581 F.2d at 506 (1978).
 55.  630 F.2d 398 (5th Cir. 1980).
 56.  Id. at 409.
 57.  In the technology transfer cases, the USEPA was attempting to use
     analyses developed for  one  type of industrial process as a basis for
     establishing an effluent guideline for other industrial processes where
     insufficient data  prevented  the Agency from developing a separate
     guideline.
 58.  See generally, C&HSugarCo. vEPA, 553 F.2d 280 (2nd Cir. 1977).
                                                               59. See generally, 514 F.2d 492 (1975) (en bane).
                                                               60. Id. at 516-517.
                                                               61. Id. at 520.
                                                               62. 501 F.2d at 1148(1983).
                                                               63. Id. at 1143. The court believed that complaint homes exhibited higher
                                                                   formaldehyde levels than non-complaint homes.
                                                               64. /
-------
           PRACTICAL USE OF RISK ASSESSMENT IN  THE
              SELECTION  OF A REMEDIAL  ALTERNATIVE
                                          KATHERINE D.  WALKER
                                          CHRISTOPHER  HAGGER
                                              Metcalf & Eddy, Inc.
                                              Boston, Massachusetts
 INTRODUCTION

  The majority of  a National Priority Listed (NPL) hazardous
 waste sites  require  the completion of a remedial  investigation/
 feasibility study (RI/FS)  as a basis for subsequent site cleanup.
 Characterization  studies  have been completed  for several NPL
 sites in sufficient detail to serve as a basis for subsequent evalua-
 tion and selection of a cleanup alternative. Other sites may require
 an RI to generate the necessary data required for the final selection
 of a remedial alternative. The Feasibility study remedial alternative
 evaluation process which includes a risk assessment is the focus of
 this paper.
  Selection  of an appropriate remedial alternative for a hazardous
 waste site ideally rests on a detailed risk assessment and on a care-
 ful analysis of the  relative impact of each alternative on public
 health, welfare and  the environment. In practice, limited data and
 time often place serious constraints on  the depth of the assessment,
 yet the need for a defensible selection of a remedial action remains.
  Concentrating on public health impact, the authors discuss Met-
 calf & Eddy's approach of focusing on the scope of a risk assess-
 ment while generating information necessary to make a defensible
 selection of a remedial alternative. The approach includes  initial
 screening of contaminants to  select  those most important  for
 assessment,  simplified models for estimating exposure and use of
 matrices to  assure consistant consideration of all risk assessment
 criteria in the selection of a final remedial alternative.

 RISK ASSESSMENT

  In recent years, several scientific papers have been published on
 the appropriate methodology to be used for evaluating the public
 health risks posed by uncontrolled hazardous waste sites. While
 most authors discuss the need  to  evaluate the various chemical,
 physical, biological and lexicological properties that affect  ex-
 posures and ultimately risks, few writers point out the major time
 and data limitations imposed upon most risk assessments of haz-
 ardous waste sites. Detailed evaluations are often not possible and
 for many sites may  not be necessary to select a remedial alterna-
 tive.
  To complete a study within a short  time frame, the  major risks
 must be identified rapidly  to screen out remedial actions that are
unwarranted and to focus on those that are most likely to effec-
tively mitigate possible impacts on public health. This approach re-
quires:
•Selecting the major contaminants of concern for the site
•Using simple exposure "models" or estimates to screen out re-
 medial actions that are unnecessary
 •Ranking the remaining alternatives according to their relative im-
 pact on public health
  Although this approach discussed below focuses on public health
 impact, the same methodology may be applied to the study of the
 other criteria used in the selection of a remedial alternative.
 Selection of Target Compounds for Assessment

  A typical uncontrolled hazardous waste site, particularly one
 that is early or pre-superfund, is the analyst's nightmare. It is likely
 to contain over  100 different compounds identified over the years
 by several investigators whose analytical techniques and objectives
 differed. Characterization of the nature and extent of site contam-
 ination is therefore often sketchy.
  The first step in conducting a "focused" risk assessment is to
 select a limited group of compounds for evaluation. This step may
 be accomplished relatively quickly by ranking the compounds iden-
 tified on-site by extent of contamination and toxicity with inclusion
 of major factors that influence their transport in the air, water or
 soil. This selection process should be carried out for each potential
 route of exposure in order to select those 10-15 compounds that are
 of concern for each transport medium.
  Often, the absence of sufficient toxicologic  or epidemiologic
 data will preclude evaluation of certain compounds from the out-
 set. The National Research Council recently estimated that only
 21% of all of the chemicals used in commerce (excluding drugs,
 pond additives and pesticides) have even minimal toxicity informa-
 tion. Data on subchronic, chronic and reproductive effects is par-
 ticularly lacking  (3-10%).'
Screening Remedial Actions

  Once the compounds have been selected for analysis, the next
step is to evaluate the relative level of risk for each route of ex-
posure. The objective of this assessment is to establish the relative
importance of each route of exposure to eliminate or minimize
further analysis of routes that have little impact.
  The relative, rather than absolute or "true", impact is the objec-
tive of the  analysis because the limitations typically imposed on
studies of this kind currently make estimates of the absolute risks
difficult if not impossible. The large uncertainties in both the con-
ceptual models and the data necessary to run them plague both
exposure and risk estimation steps. For instance, there is a growing
consensus in the scientific community that the apparent need for
groundwater modeling to study contaminant migration has general-
ly outstripped the development of  the  necessary models and
data.3-4 Even with the  best of  data, the ability of groundwater
                                                                     RISK ASSESSMENT/DECISION ANALYSIS
                                                      321

-------
modeling to predict the levels of contaminants at any given location
or time is relatively poor.' Similarly, models  used  to extrapolate
results of toxicity testing in animals at high  doses to  effects of
human exposure to low doses,  while useful,  are still largely un-
proven. The cumulative effect of these  uncertainties is to obscure
any statistical distinction between mean or best estimates of risk.
  A more suitable approach for the typical feasibility study is to
use simple models to screen  the  relative importance of various
routes of exposure under worst  case conditions. Remedial actions
involving routes of exposure that can be shown to be of little im-
portance may often be  eliminated  from detailed evaluation. A re-
cent feasibility study conducted  by M&E of an uncontrolled waste
site, the ABM Wade Property in Chester, Pennsylvania (Wade
Site), illustrates the usefulness of "screening" level models.

Wade Disposal Site
  The Wade Site covers three acres on the banks of the Delaware
River in Chester. Originally a rubber reclamation facility, the site
was used as an industrial waste storage and disposal facility in the
early 1970s. More than 4,000 drums of toxic industrial waste were
stored on the property or had their contents poured directly into
the ground. Thousands of gallons of waste were discharged directly
from tank trailers onto the ground.
  At the time of the study,  the USEPA and the Pennsylvania De-
partment of Environmental Resources  (DER)  had conducted  a
partial cleanup of the site to remove remaining barrels and trailers
of waste. A previous contractor had identified several potential re-
medial alternatives prior to any risk assessment of the site.' Of par-
ticular concern were the evaluation of the need for a slurry wall and
leachate collection system and the selection among several detailed
soil removal options. The site was known to be a playground for
local neighborhood children.
   To evaluate the potential impact of contaminated leachate on the
Delaware River and of contaminated soil on children playing on the
site, Metcalf & Eddy investigated five routes of exposure:
•Ingestion of drinking water from the Delaware River
•Ingestion of contaminated fish from the Delaware River
•Inhalation ingestion of contaminated dust from the site
•Ingestion of contaminated soil
•Inhalation of volatile contaminants emitted from the site
   The evaluation process focused on the cancer risks because of the
increased risks statistically associated with even low levels of ex-
posure. All risks were  calculated using unit risk factors (mg/kg/
day)~' developed by the USEPA Carcinogen Assessment Group.'
   To estimate the maximum level of contamination of the Dela-
ware River resulting from groundwater discharges from the Wade
Site, several worst case assumptions were used. First, the peak con-
taminant concentrations detected in groundwater on-site  were
assumed to be representative of groundwater quality throughout
the site. The maximum estimated  flow of groundwater contam-
inated at these levels  was assumed to discharge into the Delaware
River daily and to mix completely within one half of the total flow.
The calculations also assumed  no loss of contaminants  through
attenuation in soils, biodegradation or any other mechanism.
  The results of this analysis  for a  few compounds detected in
highest quantities at the site are shown in Table 1. The calculations
indicate that, even under the gross assumptions made, the final
concentrations of these chemicals in the Delaware River were like-
ly to be well below applicable drinking water criteria.
  Ingestion of contaminated fish was also a major concern because
of bioaccumulation  potential of the major  Wade  site  contami-
nants and the observation of persons fishing from the Wade prop-
erty.  To estimate  the concentration of contaminants expected in
fish, steady-state bioaccumulation factors obtained from the avail-
able literature were applied to the concentrations of contaminants
in the Delaware River calculated earlier. Increased lifetime cancer
risks from eating 6.5 g of contaminated fish were then estimated.'
  The results of this  analysis for  benzene, dichlorobenzenes  and
trichloroethylene which were all found in high concentrations in
groundwater are shown  in Table 2.  Despite bioaccumulation in fish
                                                        tissues, the contaminants were  unlikely to pose serious increased
                                                        lifetime risks of cancer.
                                                          Comparison of risks associated with contamination of the Del-
                                                        aware River with risks posed by more direct contact with contam-
                                                        inated soils suggest that the latter is potentially a more important
                                                        route of exposure. Three routes of exposure to contaminants on-
                                                        site were evaluated;  inhalation of contaminated dust (generated
                                                        during  on-site activity), direct ingestion of contaminated soil (i.e.,
                                                        by a young child) and inhalation  of organic vapors emitted from
                                                        remaining on-site waste.
                                                                                   Table 1
                                                                Impact of Wade Site on Delaware River Water Quality
                                                                                           Concent r ntIon
                                                                                           [n Delaware       Huun
                                                                              Croondwjttr   Rlvr*  (1/2      h>«lth'2)
                                                                             Concentration  flow)'''        Criteria
Con pound
Benzenr
Dlchlorobenzene
1,2 DUhloroi-thane
1,2 Dlchloropropane
To luene
1,1,1 Tr Ichloroethane
Tr Ichloroechylene
Metals
Chromium
Lead
(I) Equation used:
(C_ + i
c_ + c- - ""
u*/l
i, 100
670
6,500
7,050
12.400
21,600
5,300

1.210
2,530

'CM * CDR * **[
ug/1
0.006
0.001
0.013
0.015
0.026
0.004
0.011

O.OO2
0.005

— <0.*,\
ug/1
0.066<3>
400.00
0.94(3)

14,300
1,000
2.7(3)

50. 0<*>
so.o<4>


                                                             CM
                                                                              un    UK

                                                            where:


                                                             ' 0.01 mgd (maximum groundvmier Dux Uirough Wade Site)
                                                         QDR " 9.695 mgd (full flow of Delaware Riitn
                                                         CGM - Concentration in groundwaicr
                                                         CDR " Dala for "mbienl concentrations in DeU» are Rh er assumed to be zero
                                                         (2) Federal Rtgistrr 
-------
                            Table 2
                  Increased Lifetime Cancer Risks
             Compounds* "
                   Cancer                  ,.
                   Riskl/l   Major Assumptions1"
Ingestion of

(Adult)'
benzene           9.6 x 10 7  Dilution  in 1/2  flow of Delaware
trichloroethylene   1.4 x 10"^ Steady State Bioconcentration
dichlorobenzene    9.0 x 10"'  6.5 fish/day
                           70 kq adult
                           100» absorption  into body
Inhalation of  1,2 dichlorobenzene  2.4 x
Contaminated   di (2 ethylhexyl    2.9 x
Soil (Mult)      nhthalate)
Inqestion of
Contaminated
Soil (Child)
1,2 dichlorobenzene 8.3 x
di (2 ethylhexyl    1.3 x
   phthalate)
10~4  Dusty conditions - 10 mq/m3
10~5  1m3 breathed/hour
     8 hour exposure
     70 kg adult
     100% absorption into body

10"ij  1000 mq ingested per day
10"3  20 kq child
     100» absorbtion into body
Inhalation of benzene
Volatile     trichloroethylene
Compounds
(Adult)
                 7.3 x 10~5  8m3 breathed/day
                 2.2 x 10"5  70 kq adult
                           100» absorption through lunqs
(1) These compounds were found in highest concentrations in each medium and are assumed to be
  representations of concerns throughout the site.
(2) Risk = Increased lifetime risk of cancer. All cancer risks calculated using unit risk factors
  developed by the USEPA Carcinogen Assessment Group.

(3) These assumptions are intended to result in calculations of worst case risks.
Ranking of Remedial Alternatives; A Matrix Approach

  Use of screening models may be of little help when the remedial
alternatives are relatively similar.  The uncertainties inherent in the
risk assessment process are likely to obscure distinctions among the
alternatives. In such  cases, ranking remedial alternatives with the
help of a matrix may be simpler, faster and equally justifiable.
  The objective of the matrix is  to  ensure that all of the factors
affecting a decision are explicitly integrated into the selection pro-
cess. The alternatives fall  along the  vertical axis and the decision
factors along the other. Each alternative is ranked relative to the
others for each decision factor, total scores are tallied and final
rankings are established. The rankings are not intended to reflect
the absolute differences between the impacts of the various alterna-
tives (e.g., that one alternative is  three times worse than another).
Rather, they reflect professional judgment of the relative impact of
the various alternatives.
  The Wade Site again  serves as an example. Twelve distinct re-
medial alternatives were developed by a previous contractor. They
involved various combinations of two capping alternatives with
three soil removal options. The soil removal options involved exca-
vation to various depths dependent on "action levels" set for total
levels of contaminants in soils. The action levels were not based on
rigorous public health considerations.
  Selection of any alternative  required an implicit trade-off be-
tween the potential  for increased  short-term exposure to com-
pounds  in air during soil removal and the longer term benefits of
removal  of the more contaminated soils from the site. Rather than
attempting to model the potential exposures associated with each of
the remedial alternatives, a matrix was used.
  The matrix  used to rank the Wade Site alternatives by public
health impact is shown  in  Figure 1. Only  five of the original 13
alternatives are presented here and, of these, the soil removal Op-
tion 1 represents minimal soil excavation and Option 2 represents
maximum excavation of soil. Decision factors ranged from physical
safety hazards  and potential exposures during remedial activities
to possible occupational exposures during future development of
the site.  The final rating of the "effectiveness" of the alternative
at mitigating public health impact appears in the final column.
  The analysis suggested that an  alternative requiring a moderate
level of soil removal was ranked  more effective in mitigating im-
pacts  on public health  than both no soil  removal and extensive
soil removal options.
  Analysis of the sensitivity of the matrix results to assumptions
made in ranking the alternatives is an important final step. The
matrix implicitly gives equal weight or importance to each decis-
ion factor despite the likelihood that certain decision factors may
be considered to be more important than others or that the analyst
may have more confidence in  the ranking of alternatives for cer-
tain decision factors. Sensitivity analysis permits evaluation of the
effect of according different weight or importance to certain de-
cision factors on the final ranking of alternatives. It also helps to
reveal if the analyst's bias toward a given alternative has skewed
the results.
  For the Wade Site, a sensitivity analysis was conducted by ac-
cording greater weights first to  decision  factors related  to  short
term or acute hazards and secondly to longer term or chronic haz-
ards and  noting the effect on the final  ranking. Although not
shown here, changing the weighting factors had little effect on the
final order of alternatives, thus indicating that the result was rea-
sonably robust.
  A final ranking of alternatives that is  highly sensitive to varia-
tions in weighting factors should not be  considered a defect for
the analyst. It may simply indicate what is often  suggested  from
the outset—that a clear choice among remedial alternatives does
not exist.
  As mentioned earlier, the approach of first screening and then
ranking remedial alternatives can be used to evaluate alternatives
on the basis of the other "effectiveness" criteria that must be con-
sidered  in making  a final selection under the National  Contin-
gency Plan. The following section discusses the incorporation of
results of the individual effectiveness criteria evaluations into a
comprehensive matrix for selection of the final alternative.


FINAL SELECTION OF A REMEDIAL ALTERNATIVE
  Selection  of a cost-effective  remedial alternative must be based
on other "effectiveness" and cost criteria in addition to the public
health risks. These criteria include environmental risks, institu-
tional  issues,  implementability/reliability  issues, construction/
implementation costs and post closure, long-term  monitoring and
maintenance costs.
  M&E has adopted and modified an approach for evaluating
effectiveness and cost issues for remedial alternatives based on a
Methodology Manual prepared for the USEPA10  and a  USEPA
Guidance Document."  M&E's focused approach  is based on the
following principal components:
•Reliance exclusively on data generated during previous investiga-
 tions of the site
•Qualitative effectiveness criteria assessment and quantitative cost
 criteria assessment
•Use of matrices for criteria and alternative assessment
•Use of a matrix as a basis for selection of a cost-effective remedial
 alternative
  Metcalf & Eddy's approach to evaluating the cost-effectiveness
of remedial alternatives in a focused, timely manner is highly de-
pendent on the four principal components previously listed. The
following brief discussion will describe in more detail these com-
ponents and how they have been applied to the Wade Site as pre-
viously described.
  Generally,  a  focused feasibility  study  is conducted for a  site
where prior studies have provided sufficient data for the evaluation  .
of remedial alternatives. The Wade Site had approximately 40 re-
ports or other file documents available for the feasibility study. All
40 documents were arranged chronologically and evaluated by the
study team for applicability to the effectiveness and cost criteria.
  Even after evaluating all existing data, several assumptions may
be necessary to complete a focused feasibility study. Each assump-
tion must be valid,  realistic,  applicable and serve to focus the
remedial alternative selection  process for the particular site. The
objective in stating any assumption is twofold: (1) to reduce any
necessary assessment of site data and (2)  to provide the necessary
data for input into the alternative selection matrix.
                                                                            RISK ASSESSMENT/DECISION ANALYSIS
                                                                                                                    323

-------
              REMEDIAL
             ALTERNATIVES
                                                                                                    EFFECTIVENESS
                                                                                            TOTAL     RATINGS'
                                       10
         REMOVE ON-SITE WASTE
         GRADE PROPERTY
         REMOVE ON-SITE
         WASTE SITE CAP
         REMOVE ON-SITE WASTE
         EXCAVATE SOIL OPTION 1
         REMOVE ON-SITE WASTE
         EXCAVATE SOIL OPTION 2
          'Effectivenen ratings were ranked in revetw ol public health impact!
           10 Imott effective! - 1 I lust effective)                          Figure 1
                                      Public Health Risk Assessment-Wade Site-Chester, Pennsylvania
   Effectiveness criteria assessments are generally conducted on  a
 qualitative basis,  whereas cost criteria  assessments are quantita-
 tive. Effectiveness criteria include environmental, institutional and
 implementability which are all assessments used in the ranking of
 various  proposed remedial alternatives relative to one  another.
 For most sites, including the Wade Site, quantitative distinctions
 often cannot be made on the basis of available site data.
   Evaluation of remedial alternatives on the basis of institutional
 issues  (permit requirements, community relations, etc.) may also
 be handled on a qualitative basis for studies completed in a short
 time frame. For example, at the Wade Site, those remedial alterna-
 tives involving off-site transport and disposal of hazardous wastes
 may result in greater  institutional impacts due to permitting re-
 quirements and lesser impacts in terms of a community  relations
 program.
   Assessment of remedial alternatives based on implementability/
 reliability issues is also qualitative but is based  on a variety of fac-
 tors including the  durability of the alternative, ease of installation
 and the time needed for cleanup. Engineering judgment regarding
 proven remedial alternative technologies is critical to this successful
 assessment. For example, the site capping/waste  removal as pro-
 posed  for the Wade Site made greater use of proven technologies
 than those minimal cleanup alternatives and, therefore, were rated
 as more highly reliable.
   Cost criteria assessment  should be  handled in a quantitative
 rather than  qualitative  manner. Present worth  implementation
 costs for proposed remedial alternatives  may be calculated  based on
 standard cost estimating procedures. Post closure, long-term mon-
 itoring costs  for each alternative can also be quantified based on
 certain site specific assumptions. Quantitative  cost criteria assess-
 ment for the  Wade Site indicated important differences in over-
 all alternative costs which may have been difficult to determine if
 this assessment had been performed on a qualitative basis.
   It is important that each  assessment  of individual effectiveness
 criteria was performed by  a  different  professional as part of  a
 multi-disciplinary evaluation team. All  members of the team were
brought together to develop the final remedial alternative selection
matrix only after completion of the individual  assessments  of  the
various effectiveness criteria.  This procedure helped to eliminate
                                                         any bias toward any one particular remedial alternative in advance
                                                         of the preparation of the final selection matrix.

                                                         FINAL REMEDIAL ALTERNATIVE SELECTION: USE OF
                                                         A MATRIX
                                                           A schematic of a matrix used in the Wade Site alternative selec-
                                                         tion process is shown  in Figure 2. Effectiveness and cost criteria
                                                         are listed across the top while the alternatives are listed along the
                                                         side.
                                                           This comprehensive matrix integrates cost with assessments of
                                                         the effectiveness criteria (public health, environment, institutional,
                                                         technical) for each alternative  to aid in  the selection of the most
                                                         cost-effective remedial alternative consistent with the objectives of
                                                         the National Contingency  Plan. For each alternative, final effec-
                                                         tiveness  and effectiveness/cost  ratings have  been developed and
                                                         are presented in  the last two columns of the matrix. Final effec-
                                                         tiveness ratings are calculated by multiplying the weights accorded
                                                         a given effectiveness measure by  the individual effectiveness rat-
                                                         ings given  for each criterion. Final  cost ratings are calculated  by
                                                         multiplying the sum of the implementation and monitoring costs
                                                         by the weighting  factor. The effectiveness/cost rating for each re-
                                                         medial alternative is the division of the two separate ratings dis-
                                                         cussed above. A  higher relative rating indicates that a  particular
                                                         alternative is more cost-effective.
                                                           The weighting  factors,  which  appear in  the  top row of the
                                                         matrix, reflect the relative priority given to each effectiveness meas-
                                                         ure and  to costs. The relative priorities should be established in
                                                         advance  by the lead agency for the site. These weighting factors are
                                                         important  determinants of the overall  results of the cost-effec-
                                                         tive  matrix  analysis. The sensitivity of this analysis may  be de-
                                                         termined by changes in the assignment of weighting factors to each
                                                         effectiveness or cost criterion. The matrix easily permits this kind
                                                         of sensitivity analysis.
                                                           Figure 2 is part of  the  Wade Site Remedial Alternative Cost-
                                                         Effectiveness Matrix. The  weighting factors, which appear in the
                                                         top row  of the matrix, reflect the relative priority given to each
                                                         effectiveness measure and  cost by the USEPA's Office  of Waste
                                                         Programs Enforcement. Highest priority is given to public health
                                                         followed by the protection of the environment, implementability
324
RISK ASSESSMENT/DECISION ANALYSIS

-------
                                   REMEDIAL
                                 ALTERNATIVES
                          WEIGHTING FACTORS
                        1. NO ACTION
                        2. REMOVE ON-SITE WASTE
                         GRADE PROPERTY
                        3. REMOVE ON-SITE WASTE
                         SITE CAP
                        4. REMOVE ON-SITE WASTE

                          EXCAVATE SOIL-OPTION 1

                          SITE CAP
                        5. REMOVE ON-SITE WASTE

                          EXCAVATE SOIL-OPTION 2

                          SITE CAP
                                                            Figure 2
                             Remedial Alternative Cost-Effectiveness Matrix, Wade Site-Chester, Pennsylvania
and reliability of the alternative institutional issues and finally cost.
  The NCP dictates that the remedial alternative should be selected
on the basis of both  cost-effectiveness and the  environment. In
part, the NCP states that selection of a remedial alternative should
be based on: "the lowest cost alternative that is technologically
feasible and reliable and which effectively mitigates and minimizes
damage  to and provides adequate protection of Public Health,
Welfare, and the Environment."12
  On this basis, the no-action alternative was eliminated. Although
it has a high "effectiveness cost" rating (a function of the small
cost  of doing nothing to change existing site conditions),  it was
rated lowest in terms of effectiveness and, therefore, was not con-
sidered further. The non-soil removal alternatives (2 and 3) were
also eliminated using the same reasoning.
  The remaining contaminated soil, removal alternatives (4  and 5)
have the highest "effectiveness"  ratings, the most  "effective" of
these being alternative 5. This also had the highest effectiveness/
cost  rating of the soil-removal  options and, on this basis,  was
recommended to the USEPA for selection as the most cost-effec-
tive remedial alternative for the Wade Site.

CONCLUSIONS
  Simple models and  decision matrices can often be useful tools
for evaluating and organizing the complex conditions often found
at uncontrolled hazardous waste sites. They can assist in the selec-
tion of a cost-effective  remedial alternative within the data and time
constraints typically placed on feasibility studies.
  The public health assessment of remedial  alternatives for the
Wade  Site  in  Chester,  Pennsylvania illustrates  the  potential
strength  of simple, worst case models. Prediction of the impact of
contaminated groundwater on the Delaware River depended  not on
a detailed, time-consuming evaluation of the various geological,
chemical  and biological factors affecting pollutant  transport in
groundwater, but on rapid simple estimates of flow through the
site into the river. In this case, a more detailed evaluation was un-
warranted. The simple estimates that were made permitted the
elimination of a complex leachate collection and treatment system
from further consideration.
  Decision matrices help formalize the decision-making process.
They  are intended to  assure that all  of the factors affecting the
selection of a remedial alternative are integrated into the evalua-
tion. Matrices  are therefore useful at each stage in the selection
process from evaluation of alternatives according to each "effec-
tiveness"  criterion to selection of a final cost-effective remedial
alternative.
  As any other  tools, simple  screening  models and decision
matrices have their limitations; they may not be appropriate for
all situations, nor will they always lead to a clear choice of a re-
medial alternative. However, by helping to limit the number al-
ternatives and to  consistently consider all factors affecting a de-
cision, they help assure a more defensible selection.


REFERENCES

 1. National Research Council, Toxicity Testing; Strategies to Determine
   Needs  and Priorities. National  Academy Press,  Washington, D.C.
   1984.
 2. Fiering, M. and Wilson, R., "Attempts to Establish Risks by Anal-
   ogy," Risk Analysis, 3, Sept. 1983.
 3. Anderson, M.P., "Using Models to Simulate the  Movement of Con-
   taminants  through  Groundwater Flow Systems". Crit. Rev. in  En-
   viron. Control, 9, Nov., 1979.
 4. Konikow,  L.F., "Role of Numerical Simulation in Analysis  of
   Groundwater  Quality Problems. The  Science of  the Total Environ-
   ment, 21, 1981.
 5. Anderson,  M.P., "Groundwater  Modeling-The  Emperor Has No
   Clothes",  Editorial adapted from  a talk delivered to  the Minnesota
   Groundwater Association, May, 1983.
 6. Weston, R.F., Inc., Site Characterization Activities on the Wade Prop-
   erty, Chester, PA.  Final Draft Report to Pennsylvania Department
   of Environmental Resources, Nov., 1983.
 7. Anderson,  R.A. and the  Carcinogen Assessment Group, USEPA,
   "Quantitative  Approaches in Use to Assess Cancer Risk," Risk
   Analysis, 3, Dec., 1983.
 8. Mitteleman, A., Toxicologist, Office of Waste  Programs Enforce-
   ment, USEPA, Personal Communication, July 11,1984.
 9. Federal Register, "Water Quality Criteria Documents; Availability".
   45, No. 231: 79318-79379, Nov. 28, 1980.
10. Radian Corp., Evaluating  Cost-Effectiveness of Remedial Actions at
   Uncontrolled Hazardous Waste Sites, prepared for Solid and Hazard-
   ous  Waste Research  Division, Municipal  Environmental Research
   Laboratory, USEPA, Cincinnati, OH,  Jan., 1983.
11. JRB Assoc., Superfund Feasibility Study Guidance Document, pre-
   pared  for the  Municipal Environmental  Research  Laboratory,
   USEPA, Cincinnati, OH, Aug., 1983.
12. Federal Register, 47, No. 137: 31217, July 16,1984.
                                                                            RISK ASSESSMENT/DECISION ANALYSIS
                                                            325

-------
              CLEANUP  COST ALLOCATION (CCA)  MODEL
                                             RICHARD B.  ADAMS
                                          PHILIP M. ZIMMERMAN
                                             G&E Engineering, Inc.
                                             Baton  Rouge,  Louisiana
                                      DONALD  D. ROSEBROOK, Ph.D.
                                            ENDO  Environment,  Inc.
                                             Baton  Rouge,  Louisiana
INTRODUCTION
  There are two major issues to be settled at most Superfund sites
and at any waste cleanup scene where more than one party is in-
volved: (1) agreement with regulatory officials as to the cleanup
response required and (2) agreement among involved industries as
to the apportionment of cleanup costs. The former issue has been
given much attention and while difficult to address,  has been sub-
ject to rational technical analyses. The latter issue is quite different
and typically has not been approached in a rational or  consistent
manner.
  In  this  paper,  the authors  present  a cost apportionment
mechanism which has been applied to a major Superfund site and is
being considered for use at others. The mechanism is termed  the
Cleanup Cost Allocation  (CCA) Model.1'2 Development  of  the
model began in January, 1983, when one potentially responsible
party (PRP) involved in the PetroProcessors, Inc.  (PP1) aban-
doned waste site near Baton Rouge, Louisiana took exception to its
proposed "fair-share" of cleanup costs.
  The authors were retained by this industry to develop a rational
cleanup cost allocation model which could be applied to  essentially
any waste site. The model was applied to the PPI site from May
through November, 1983.'
NEED FOR THE MODEL
  The model was developed to fit the perceived need to account for
substantial differences in the nature of wastes. The following exam-
ple underscores the urgency for the acceptance of a waste-related
apportionment scheme for allocating cleanup cost. Consider a site
(with  $10 million cleanup costs) consisting of 100,000  tons of
relatively innocuous products from 20 generators and 1000 tons of
known carcinogenic waste from a  21st generator.  On  a  strict
volume basis, the toxic waste generator would pay only $100,000.
More  realistically, however, the first 20 generators should pay the
cost equivalent of having to close a solid waste landfill with all re-
maining costs being picked up by the toxic waste generator.
  Currently,  neither government nor  industry  has  formal
guidelines for handling cleanup cost allocation. Indeed, the issue of
joint and several liability  precludes government commitment to
such guidelines.  Without an organized approach toward  accoun-
ting for waste properties and  tendencies, cleanup costs for aban-
doned waste sites would be allocated among responsible parties
based  on volume alone,  depth-of-pocket or at best those two fac-
tors would  be considered along with an unorganized estimate of
harm done.
                                                    CONSIDERATIONS FOR COST APPORTIONMENT
                                                      Thus, it was clear from the beginning that the model must incor-
                                                    porate waste-related factors impacting cleanup cost. Here, it is im-
                                                    portant to differentiate between the factors which impact the cost
                                                    to cleanup a waste site and the factors which serve as a reasonable
                                                    basis  for cost apportionment  among involved parties. Principal
                                                    determinants of a waste site cleanup effort are listed below:
                                                    •Nature and quantity of wastes
                                                    •Site  conditions (e.g., topography,  geology, hydrology and  cli-
                                                     matology)
                                                    •Disposition of wastes (co-mingled, distinctly separate or a com-
                                                     bination of the two)
                                                    •Interaction of wastes with site (surface water, soils/rock, ground-
                                                     water, air and living environment) and with other wastes
                                                    •Criteria for cleanup (e.g., regulatory requirements and allowable
                                                     residual contaminant levels)
                                                    •Closure and post-closure response adopted
                                                    •Public interest
                                                      Each of the above items impacts the cost to clean up a waste site.
                                                    However,  the only reasonable differentiating variable  for appor-
                                                    tioning cost among involved parties is the waste itself. A given set
                                                    of wastes deposited in a clay environment versus a sandy environ-
                                                    ment  will, other things being equal, require less expense to clean
                                                    up. However, apportionment  of the cleanup cost  in  either case
                                                    should be according to the wastes involved—regardless of the total
                                                    costs.
                                                      What are the waste factors that most significantly impact the cost
                                                    of cleanup and how should they be weighted? The CCA model  was
                                                    developed to answer these questions.

                                                    MODEL DESCRIPTION
                                                      Basically, the CCA model is a mechanism for  equitable alloca-
                                                    tion of "site" cleanup costs. Equitable cost allocation is defined as
                                                    apportionment of cleanup expense among involved parties accor-
                                                    ding to  the impact their wastes have on total cleanup cost. Con-
                                                    siderations for the model included: prior cleanup cost allocation ef-
                                                    forts; existing hazard ranking systems; available cleanup cost infor-
                                                    mation for on-site and off-site disposal and treatment; and cleanup
                                                    methodology requirements.
                                                      The model is based on three fundamental criteria: (1) different
                                                    wastes can pose substantially different risks to human health  and
                                                    the environment,  (2) cleanup methodology and costs reflect con-
                                                    cern for—and characteristics of—the wastes involved and (3) waste
                                                    site cleanup typically involves not only the waste itself but also off-
                                                    site contaminant migration.
326
COST OF CLEANUP

-------
  "Site" is defined in the basic model as the land area physically
used for treatment, storage and/or disposal of wastes, plus an ar-
bitrary 100 ft lateral buffer zone (beyond the limits of the waste
management  land area) and a 10 ft vertical buffer zone (below the
base of waste units).  The inclusion of lateral and vertical buffer
zones is a subjective and conservative effort to accommodate ill-
defined waste deposition areas. However, "off-site" cleanup costs
for highly mobile waste materials are still segregated from the less
mobile wastes. Those  highly mobile wastes will require separate
cleanup cost  apportionment incorporating site-specific data.
 MODEL CONSTRUCTION

   The  CCA  model rationale incorporates apportionment  of
 cleanup costs according to assignments given to weighting factors
 for individual wastes. These factors comprise three major com-
 ponents listed below: (1) waste quantity, (2) biological (toxic) fac-
 tors and (3) physical factors:
 •Waste Quantity
 •Biological (Toxic) Factors
  Acute toxicity
  Carcinogenicity
  Mutagenicity
  Teratogenicity
  Reproduction
  Subchronic and other toxic effects
 •Physical Factors
  Consistency
  Mobility
  Activity
  Persistence
  Ignitability
  Corrosivity
  Reactivity
   Scores are assigned to the above factors for each waste. Scoring
 follows rigorous compilation and evaluation of waste-related data.
 The individual waste cleanup cost allocation factors are then com-
 puted as follows:
   Individual Waste Cleanup Cost Allocation Factor =
      (AxBxC)/B(AxBxC),                                  (1)
   Where for any given waste:
       A = Waste quantity (common units for all wastes)
       B = Highest score assigned to the array of Biological
             Factors
       C = Sum total of scores assigned to the array of Physical
             Factors
   The cleanup cost allocation factor for each party is then the sum
 of its individual  waste cleanup cost allocation factors.
   It is important to recognize that: (1) the standard definitions of
 toxicity do not allow sufficient discrimination to truly evaluate tox-
 icity as  a cost impact  factor and arbitrary scoring systems have to
 be established; (2) the probability that certain waste-waste interac-
 tions might occur must be realistic and allow discrimination be-
 tween wastes; and (3) as often as possible, guidelines for applying
 the model must be concrete.
   The three major model components are discussed below.
 Quantity

   Waste quantity is still the dominant consideration. A PRP's par-
 ticipation in a cost allocation program  is obviously based on the
 presence of his waste at  the site. The total weight or volume at-
 tributed to the waste is Factor A in Equation 1.
 Biological (Toxic) Factors
  The potential of a waste to do harm to human health or the en-
 vironment  directly impacts  the effort  and cost of waste site
 cleanup. Cleanup considerations which reflect  concern  for this
 potential include:
•Extent of cleanup effort
•Cleanup methodology
•Long-term monitoring and testing
•Public interest
  The difference in cleanup effort required for a non-hazardous or
low hazard project and that required for a high hazard project is
substantial.  The difference may range from selected (or limited)
waste removal and/or site capping in the low hazard case to exten-
sive material removal,  subsequent treatment and disposal of con-
taminated groundwater,  construction of containment facilities,
etc., in the high hazard case.
  The cleanup protocol necessary to safely handle wastes and ac-
complish  assigned cleanup  work will generally  be much  more
rigorous for a high hazard site than for a non-hazardous site. The
degree of worker protection and the corresponding worker efficien-
cy can vary  substantially.
  The number of monitoring locations and the sophistication and
expense of well installation and analytical testing will invariably
reflect the potential hazard posed by the site.
  Experience has shown that public sentiment and non-technical
involvement can have a significant impact on the level of effort in-
volved in remedial activities.
  The above considerations led to the assignment of weighting fac-
tors of 1, 3, 5 and 7 for each of the subfactors contributing to the
Biological Factor.  However, if the subfactors are considered to be
interactive in any way, the weight given to the Biological Factor is
disproportionate. Therefore, only the subfactor with the highest
score is considered and this weight is assigned as the Biological Fac-
tor (B) in  Equation 1.
Physical Factors

  Wastes typically have differing physical characteristics which im-
pact cleanup costs differently. The Physical Factor C in Equation 1
is the sum of weightings given to seven individual physical com-
ponents:

  Physical Factor (C) = (CON + M+A + P + I + COR + R)     (2)
  Where:
     CON =  Consistency
      M  =  Mobility
       A  =  Activity*
       P  =  Persistence
        I  =  Ignitability
     COR  =  Corrosivity
       R  =  Reactivity
     * Activity = (Am+ Ar + Ae)/6                           (3)
  Where:
   Am = A + B + C                                        (4)
     Ar = D + E                                            (5)
     Ae = F + G                                            (6)
      A = Response to Water
      B = Response to Chlorinated Solvents
      C = Response to Hydrocarbon Solvents
      D = Reactivity w/Organics
      E = Reactivity w/Inorganics
      F = Solvent Strength
      G = Absorptive Capacity

Consistency

  Costs for solidification of wastes vary by a factor of 3 between
solids and liquids. In the case where treatment and discharge of a
liquid aqueous waste is selected as an alternative to solidification,
the cost differential between a solid waste and a liquid is still a fac-
tor of 3. Thus, the weighting for consistency becomes 1 for solids, 2
for sludges and 3 for liquids.
                                                                                                 COST OF CLEANUP
                                                          327

-------
Mobility

  Costs associated with the mobility of wastes reflect the potential
of a given waste to move (rather than actual measured movement
of the wastes). It was necessary to limit the concept of mobility to a
general potential to move in order to keep the model site indepen-
dent. Briefly, the weighting  for  mobility is 1  for solids,  2 for
sludges, 3 for permeants similar to water and 4 for permeants which
tend to  move faster than water. The mobility factor does not ad-
dress the cost of cleaning up off-site contamination resulting from
waste migration.  Application of the model to off-site contamina-
tion cleanup requires site-specific information.
Activity

   Since it probably is not fair to allocate costs based only on the
possibility that a certain interaction  may potentially occur  if the
right materials come together under the right conditions, the effects
on the allocation  factor must be tempered. This has been achieved
in the activity term.
   Within the  physical factor "activity", three sets of subfactors
have been identified. These subfactors (a through g) comprise a
mobility term—Am;  a  reactivity  term—Ar;   and  an effects
term—Ae.
   Weightings of 1, 2 and 3 are assigned to subfactors a through f
and a weighting of - 1 or 0 is assigned to subfactor g.  A conser-
vative approach dictates that  terms a through f are all positive,
presuming that if an interaction does occur then the results will be
to enhance either the mobility or the toxicity of something at the
waste site. Only term g can affect a reduction in the activity factor
term.

Persistence
   Persistence weightings of 1 through 4 are similar to those used by
the USEPA in the "Mitre Model". The degree  of persistence af-
fects the  time-dependent post-cleanup requirements  for  site
monitoring.
Ignitability, Corrosivity and Reactivity
   Weightings  for ignitability (0 through 4),  corrosivity (0 and 1)
and reactivity (0 through 4)  are similar to widely used scales in
USEPA regulations. Ignitability  impacts normal cleanup  costs
because of a greatly increased handling problem encountered when
handling ignitable wastes. Scoring corrosivity  and reactivity is
reserved for containerized waste.

STRENGTHS OF THE CCA MODEL
   The CCA model embodies  several distinct  features and affords
negotiating parties an independent  cost allocation mechanism:
•It can be applied to essentially any waste site
•It can be applied prior to site investigation  for apportionment of
 on-site cleanup cost; application for off-site .cleanup cost alloca-
 tion requires site-specific information
•Input data are waste-related,  not site-related
•It can be readily updated as  new information becomes available
•It provides incentive for full disclosure of waste characterization
 data
  The CCA model is a viable mechanism for apportioning cleanup
costs for waste sites. At a minimum, it provides a reasonable basis
for negotiations among PRPs.

USE OF THE CCA MODEL

  Applying the CCA model requires information  about the wastes
present at a site. The information can be either general and limited
or comprehensive. Obviously, the more detailed the waste  data
base the more accurate the. model output. Specific input for the
model includes the following:
•Identification of the participants in cleanup cost apportionment
 (potentially responsible parties, PRPs)
•Description  of individual  wastes  sufficient  for characterization
 and evaluation; wastes which are unknown or inadequately de-
                                                         scribed are scored conservatively (high) by the model,  thus pre-
                                                         cluding undcr-statement or possible waste impact
                                                        •Determination of quantities of individual waste types attributable
                                                         to each PRP
                                                        •Identification of consistency of each  waste at the time  of its de-
                                                         position at the site; wastes with unknown consistencies are con-
                                                         servatively considered to be liquids
                                                        •Determination of locations of deposited wastes, migration plumes
                                                         (surface and subsurface) and other evidence which confirms ap-
                                                         propriateness of subdividing cleanup  cost activities among fewer
                                                         than all PRPs; in  the absence  of such information wastes are
                                                         assumed to be co-mingled.
                                                         To the extent possible, the above information should be substan-
                                                        tiated by shipping documents, purchase orders, invoices, analytical
                                                        test results, knowledgeable testimony,  etc.
                                                          Effective application of the CCA Model presupposes that input
                                                        data (waste quantities,  biological factor  assignments and physical
                                                        factor assignments) have been properly developed and, where ap-
                                                        propriate, reasonable assumptions made. In particular, evaluation
                                                        and scoring of waste-related biological  (toxic) factors and physical
                                                        factors must be accomplished by persons with the technical back-
                                                        ground to make realistic, defendable judgments.
                                                        MODEL APPLICATION

                                                          Generally, there will be both a static and a dynamic phase to the
                                                        cleanup of a waste site. The static phase is concerned with the site
                                                        proper (as defined previously). The dynamic phase is concerned
                                                        with efforts to mitigate contamination which has migrated (surface
                                                        or subsurface) away from the site. The CCA model effectively ad-
                                                        dresses the static phase, cost  apportionment given only knowledge
                                                        of the wastes.  Application of the model to the dynamic phase re-
                                                        quires, in addition to input waste information, sufficient  site-
                                                        specific  information  to  establish what  wastes  (or  waste  consti-
                                                        tuents) have moved off-site, the extent of their movement and their
                                                        impact on cleanup cost for areas beyond "site" limits.
                                                        EXAMPLES OF THE MODEL'S APPLICATION
                                                          Use of the model for apportionment of static phase cleanup cost
                                                        is illustrated by the three cases below, which reflect increasingly
                                                        more complex (and realistic) waste-related conditions. It should be
                                                        assumed that in each case the three PRPs have numerous types of
                                                        wastes comprising the quantities noted below.
                                                            Company I     — 1000 tons
                                                            Company II    — 2000 tons
                                                            Company III   — 3000 tons
                                                        Case 1:
                                                          All wastes have proportionately the same biological and physical
                                                        effects for each PRP
                                                        Case 2:
                                                          Waste quantities and physical characteristics are the same as for
                                                        Case 1, but waste biological (toxic) effects differ for each PRP.
                                                        Case 3:
                                                          Waste quantities are the same as for Case 1, but waste biological
                                                        (toxic) effects and physical  characteristics differ for each PRP.
                                                          The over-simplified model application for these three  cases is
                                                        shown in Table 1; a summary of the allocation percentages for each
                                                        case is presented below.

                                                                                COST ALLOCATION (%)

                                                          Case          Co. I   Co. II   Co. Ill
                                                            1             17       33      50
                                                            2             35       50      15
                                                            3             50       44       6


                                                          These simplified cases illustrate the significant impact on cleanup
                                                        cost apportionment recognized  by the CCA model as pertinent
328
COST OF CLEANUP

-------
muie 1
Example Cases

COMPANY
I
II
III

Cleanup

COMPANY
I
II
III

Cleanup

COMPANY
1

(A)
WASTE
QUANTITY
1000
2000
3000


CASE 1
(B) (C)
BIOLOGICAL PHYSICAL
FACTOR FACTOR
5
5
5

Cost Apportionment:

(A)
WASTE
QUANTITY
1000
2000
3000


8
8
B
(E)
Company I
Company II
Company III
CASE 2
(B) (C)
BIOLOGICAL PHYSICAL
FACTOR FACTOR
7
5
1

Cost Apportionment:

(A)
WASTE
QUANTITY
1000
2000
3000

8
8
8
(E)
Company I
Company II
Company III
CASE 3
(B) (C)
BIOLOGICAL PHYSICAL
FACTOR FACTOR
7
5
1
13
8
4

(D)
AxBxC
40,000
80,000
120,000
240,000
- 17Z
- 33Z
- 50Z

(D)
AxBxC
56,000
80,000
24,000
160,000
- 35Z
- 50Z
- 15Z

(D)
AxBxC
91,000
80,000
12,000

COST ALLOC.
FACTOR
0.17
0.33
0.50
1.00


COST ALLOC.
FACTOR
0.35
0.50
0.15
1.00


COST ALLOC.
FACTOR
0.50
0 .44
0.06
                                (E)
                                       183,000
                                                    1.00
Cleanup Cost  Apportlonraent:
Company I    -  502
Company H   -  44%
Company HI  -  06%
waste-related factors are varied. A more comprehensive example of
how to use the CCA model is presented in Reference 1.
  Discussions of the general model application to the PPI site have
been presented in previous papers.3-4 The PPI Superfund site (ac-
tually  two sites)  encompassed what would be considered a wide
range of waste materials and site conditions.
SPECIAL CONSIDERATIONS
  The weighting of the CCA model factors reflects typical wastes
which may  be found  at abandoned  wastes  sites.  There will be
special cases in which the wastes are inadequately handled by the
model. Such cases will involve wastes which have extreme toxicity
and  high  persistence, including dioxin  (2, 3, 7, 8  tetrachlorodi-
benzo-p-dioxin) and radioactive isotopes.
  The problems presented by these types of wastes are two-fold:
•The waste  must be removed  to very low levels of background
 contamination; cleaning a site to very  low background levels
 usually  means that inordinately large amounts of background
 soil and groundwater must be removed and/or treated
•High levels of personnel  health and safety protection  must be
 maintained at all times in all areas
  These problems can add to the cost  in a very dramatic way.
  Since the weighting factor for toxicity reflects the cost of the ex-
tent  of cleanup effort, special cleanup methodology, long-term
monitoring and testing and public interest considerations,  this fac-
tor must be  modified or overriden for  special considerations.
  There may be several ways to approach this problem. However,
the most practical would appear to be to apply the model  as if the
problem  waste were not present.  The  absolute cost  of  cleanup
would be  calculated as if the problem waste were not present and
again for  the total situation as it really exists.  The costs for the
hypothetical closure would be allocated according to the model and
the excess costs would be allocated to the generator  who disposed
of the problem waste.

CONCLUSIONS
  Historically, allocation of cleanup costs for waste sites has been a
source of confusion and frustration. The  CCA model is  a viable
mechanism  for apportioning cleanup costs for waste  sites. The
model considers not only volume but  also the risks  posed by and
the physical characteristics of  individual  wastes as they impact
cleanup costs.

REFERENCES
1. Adams,  R.B., Zimmerman, P. and Rosebrook, D., Abandoned Waste
  Site Cleanup Cost Allocation (CCA) Model, G&E Engineering,  Inc.,
  Baton Rouge, LA, Nov., 1983.
2. Adams,  R.B., Zimmerman,  P., Rosebrook, D. and Parent, R., "Aban-
  doned Waste  Site  Cleanup Cost Allocation (CCA) Model", Poster
  Paper, Society of Environmental Toxicology and Chemistry (SETAC)
  Conference, Nov., 1983, Crystal City, VA.
3. Adams,  R.B.  and Rosebrook, D., "Rational Allocation of  Cleanup
  Cost for Superfund Sites", Chemical  &  Radiation Waste Litigation
  Reporter, 7, Nos. 2 & 3, 1984, 165-170.
4. Rosebrook, D.D.  and Abrams, R.,  "Fair Share Apportioning of
  Hazardous Waste Site Cleanup Costs", Proc. of the Can-Am Chemical
  Congress, Montreal, Canada, June, 1984.
                                                                                              COST OF CLEANUP
                                                                                                                          329

-------
              REMEDIAL COST ESTIMATION SYSTEM  FOR
                                        SUPERFUND SITES

                                          KENNETH T.  WISE, Ph.D.
                                         Putnam,  Hayes & Bartlett, Inc.
                                            Cambridge,  Massachusetts
                                                PAUL AMMANN
                                             Boxford, Massachusetts
 INTRODUCTION

   The work described in this paper is part of an ongoing project
 sponsored by the Office  of  the Comptroller  and the  Office of
 Emergency and Remedial Response of the USEPA. The primary
 purpose of this project is to develop an estimate of the total costs of
 remedial actions at sites that are currently on or might be added to
 the  National  Priorities  List (Superfund sites)  as an  aid  to
 budgeting, planning and policy development. The first phase of the
 project took place  from September to December 1983 and is
 described in this paper.
   When  this project started, remedial activity had only been com-
 pleted at a small number of the over 500 Superfund sites. Further-
 more, only a small number of feasibility studies  that prescribed
 alternatives and estimated potential remedial costs had been com-
 pleted. Thus the available  database on  actual and  estimated
 remedial costs represented a small percentage of the Superfund
 sites and did not by itself allow one to draw strong conclusions con-
 cerning the average cost per site across the entire National Priorities
 List (NPL).
   Consequently,  it was necessary to develop additional informa-
 tion that could be used to estimate the costs at the majority of NPL
 sites for which credible and consistent estimates did not currently
 exist. However, it was also clear that some simplification of the
 problem  was necessary since detailed cost estimates could not be
 developed for all sites, or even for one site, within the scope of this
 effort.
   In many such cases, the simplification takes the form of a model
 plant or model site that can be used to represent a large number of
 similar situations and thus facilitate  extrapolation. The authors'
judgment was that in the  case of NPL sites the variability of the
type and  scope of problems among sites would make it difficult to
identify or create one or more model sites that would be represen-
tative of a large number of similar sites.
  For example, a large number of sites are landfills. Relative to
other categories such  as chemical plants or manufacturing plants,
landfills probably have a low degree of variation in cost.  However,
there are substantial differences among landfills: size, topography,
extent of subsurface contamination, leachate formation,  proximity
to houses and wells and  degree of groundwater  contamination.
This variability makes it very difficult  to identify a particular land-
fill that would on average be  representative. Furthermore, even if
enough information  about  other landfills on  the NPL  were
available so that a representative landfill could be designed, there
would be no assurance that it would be representative of landfills
that  might be added to the NPL in the future.
                                                        Rather  than try to design representative sites, the authors at-
                                                      tempted  to  deal directly with  the diversity in  problems among
                                                      Superfund sites. Each site was viewed as being composed of one or
                                                      more attributes (e.g., drums, contaminated soil, groundwater con-
                                                      tamination)  that in most cases would require specific remedial ac-
                                                      tions (e.g., drum removal, soil removal and capping, groundwater
                                                      treatment). The costs at a specific site could then be estimated by a
                                                      three-step process in which (1) data concerning the problems at the
                                                      site  were collected, (2)  the remedial actions likely to be taken in
                                                      response to these problems were determined and (3) cost estimates
                                                      were developed for each remedial action using functions that incor-
                                                      porated the major cost-creating factors (e.g., the number of drums
                                                      and  the distance to a secure landfill).
                                                        By focusing on  a relatively small number of attributes that give
                                                      rise  to the major remedial costs and by developing cost functions
                                                      that could be used repetitively, a large number of reasonably ac-
                                                      curate and unbiased site-specific cost estimates could be developed
                                                      at relatively  low cost. This would enable the wide diversity in site
                                                      characteristics to be considered directly and avoid the need to ex-
                                                      trapolate to  the entire NPL from a small  sample. This approach
                                                      would also establish a framework for considering additional sites
                                                      that could be added to the NPL, for evaluating expeditiously the
                                                      cost implications of the additional site information developed as
                                                      site  studies  progressed  and for considering the cost impacts of
                                                      changes in USEPA policies regarding acceptable remedial actions.

                                                      MAKING THE APPROACH OPERATIONAL

                                                        To make  this approach  work,  a  set of decision criteria  was
                                                      developed to apply to aspects of the problem at each site to deter-
                                                      mine the types of remedial actions that would generally be applied.
                                                      Functions were also developed that would use site data to estimate
                                                      the costs of  applying a particular type of remedial action. Finally,
                                                      the data required by the methodology were collected through direct
                                                      contact with the  USEPA regional personnel most familiar  with
                                                      each site.
                                                        Once data had been collected at each site, the costing procedure
                                                      involved the following actions:
                                                      •Project members reviewed the data and determined what remedial
                                                       actions would be likely at each site. The selection  was usually
                                                       based on the decision  criteria but was often influenced by the
                                                       judgment of the USEPA site manager or the judgment of project
                                                       personnel based on discussions with the site manager.
                                                      •Cost estimates were generated for each remedial action to be taken
                                                       at  a site using the cost functions.  For sites  where engineering
                                                       studies or actual costs for  remedial  actions were available, these
                                                       were used instead of the estimates.
330
COST OF CLEANUP

-------
•Cost estimates were reviewed to ensure reasonableness. In many
 cases, USEPA site managers had good estimates of what costs
 would be, and our estimated costs were compared to these to en-
 sure that our approach was reasonable. When this occurred, the
 review  process  identified costs  for particular remedial  actions
 that seemed extremely high.  In  such cases, these costs were re-
 viewed to see if similarly effective but lower cost alternatives (e.g.,
 containment and capping instead of massive soil removal) were
 available.  If so, costs were recalculated.
DECISION CRITERIA

  To implement the costing system, it was necessary to have criteria
or rules that could be used to decide what remedial actions would
be likely to occur given the nature of the problems at a site. To
develop these rules, the authors first discussed the situations at a
number of Superfund sites with regional  USEPA site managers.
For many of these sites, specific remedial actions had been taken or
could be anticipated with some confidence. The problems and ex-
pected remedial actions at these sites generally followed patterns
that were consistent with a preliminary set of decision criteria that
were developed.
  The criteria have undergone some revision as a result of cleanup
experience gained in several USEPA regions and will be further
refined  both as additional  site information is  collected  and as
USEPA policy with regard to problems such as groundwater treat-
ment or handling of contaminated soil evolves. The criteria in each
problem area that were found to be the most generally applicable in
late  1983 are discussed in the following sections.

Drums
•If above ground drums or barrels exist, then remove them to an
 off-site hazardous waste  disposal  site (HWDS)
•If buried drums exist near the surface, then excavate and remove
 them to an off-site HWDS unless they are in a landfill that will
 be capped.
•If buried drums exist well below surface, then cap the site.
Tanks
•If tanks containing liquids or sludges  exist above ground, then
 drain contents and remove to HWDS; then decontaminate  and
 dismantle tanks for salvage or disposal.
•If tanks containing liquids or sludges  exist below ground, then
 drain contents and cap area.
Lagoons
•If lagoons  contain organics  in liquid phase, then drain liquids
 and transport to HWDS.
•If  lagoons  contain aqueous solution,  then  either  drain  and
 transport to treatment or disposal site, or drain, treat on-site and
 discharge to municipal sewage treatment, depending upon which
 is less  expensive.
•If lagoons contain sludges with volatile or highly soluble chemi-
 cals, then excavate the sludge and transport to HWDS.
•If lagoons contain other  sludges,  then cover and cap.
•All lagoons are backfilled and covered with an impermeable bar-
 rier or cap.

Contaminated Soil
•If surface soil is highly contaminated with PCBs, dioxin, pesti-
 cides or other highly toxic or carcinogenic substances, then ex-
 cavate soil, transport to HWDS and cap area.
•If highly contaminated hot spots  exist, then excavate soil, trans-
 port to HWDS and cap area.
•If other types of contaminated surface soil exist, then cap area.

Buildings and Equipment
•If contaminated buildings or equipment exist, then decontaminate
 and dismantle for scrap or salvage.
Leachate

•If leachate is being produced from a landfill, dump or other site,
 then implement a collection and contamination removal system.
Surface Water Diversion
•If the topography of a site is such that off-site water would run
 through a contaminated area, then implement upgradient surface
 water diversion.

Capping

•If contaminated soil exists with which people or animals could
 come in contact, then cap with clay.
•If contamination exists on or  below the  surface and threatens
 groundwater and if the contaminant is a volatile organic or highly
 soluble chemical, then cap with clay and plastic  liner; otherwise
 cap with clay.
•If contamination exists in dirt roadways or parking lots, cap with
 clay and asphalt.
•If contaminated surface  runoff to  off-site  areas  or surface
 waters exists, then cap site with clay.
Fencing

•If public access to a site exists,  then fence entire  site.
On-Site Groundwater Treatment
•If groundwater under a  site is contaminated and the natural
 groundwater flow creates a potential or immediate threat to drink-
 ing water for which there are no economic alternatives or to other
 surface waters used for drinking, recreation or commercial pur-
 poses, then  treat groundwater  on site at the rate of 5,000 gal/
 day/acre  of contaminated area. Various  probabilities are as-
 signed to such  treatment depending upon whether groundwater
 contamination is confirmed, likely, possible or unlikely.
Off-Site Groundwater Treatment
•If contaminated groundwater has migrated off-site  and the na-
 tural groundwater flow creates a potential or immediate threat
 to drinking water for which there are no economic alternatives or
 to other surface  waters used  for drinking, recreation or com-
 mercial purposes, then treat groundwater on-site at  the rate of
 5,000 gal/day/acre of contaminated area.
Well Problems
•If public wells are  contaminated or  imminently threatened  by
 contaminated groundwater, then  either relocate  the well field if
 an alternative well field site is  close  enough to be economic or
 install a water treatment plant.
•If private wells are contaminated or  imminently threatened  by
 contaminated groundwater, then  either relocate  wells if cost  ef-
 fective, hook up homes or industries to municipal supplies if cost
 effective or  treat private wells.
Slurry Walls
•If groundwater would run through contaminated subsurface
 areas of a site, then build a slurry wall to divert  groundwater or
 contain contamination if such construction is possible.
•If on-site groundwater treatment  is chosen as a remedial action,
 then build a slurry wall if such construction is possible unless the
 threat of groundwater  contamination is  sufficiently distant in
 time that treatment without a slurry wall is adequate.
COST EQUATIONS

  Once the decisions have been made on likely remedies, the cost
functions are  applied  to generate  estimates for  the capital,
operating and maintenance costs associated with each remedial ac-
tion. For this paper, costs were  developed for the most common
remedial activities based on information available from the follow-
ing sources:
                                                                                               COST OF CLEANUP
                                                         331

-------
•Site feasibility studies
•Vendor quotations
•Contractor bids
•Actual remedial construction costs
•Publications on hazardous waste treatment projects
•In-house engineering and cost data
  For  some remedial actions, costs are  reasonably  well known
while for others the lack of data and actual site experience makes
the development of precise functions more difficult. The develop-
ment of the cost functions is continuing as part of an ongoing ef-
fort.  Without going into great  detail concerning  the  specific
numbers used, the following sections describe the nature of the
functions used in this system.
Drums
   Drums containing hazardous chemicals or materials are sampled,
staged, overpacked  when necessary and  transported to an ap-
propriate site  for disposal. The cost elements included in the cost
function are the number of drums, the percentage requiring over-
packing, the contents of the drums and the transport distance.

Tanks
   At some sites, hazardous chemicals and  materials have  been
stored  in tanks  usually located above ground. For all sites it is
assumed that  the hazardous contents are  removed from the tanks
and transported to a hazardous disposal or treatment site. The cost
include: (1) removal from tanks to tank trucks, (2) haulage to
disposal site and (3) disposal. The costs depend on the quantities
and composition of the tank contents, the distance to a disposal site
and the costs for disposal (including  incineration). It is assumed
that the tanks are decontaminated, dismantled and removed  from
the site if they are on the surface. Only in certain instances are
underground tanks excavated and removed from the site.

Lagoons, Ponds and Pits

   At many sites, aqueous  solutions, solvents, oils,  acids and other
chemicals that were  contaminated with hazardous materials  were
collected and stored  in lagoons, ponds or pits.  For this study, it is
assumed that the organics are removed to hazardous waste disposal
sites.
   The quantity of the aqueous phase  is important; small  volumes
are assumed to be removed to approved disposal sites, but very
large volumes could be treated on-site. Sludge at the bottom of
lagoons, ponds and pits is excavated and removed from the site. All
areas are assumed to be backfilled and capped to prevent infiltra-
tion of precipitation.
   The  removal  of solutions and  sludges is  expensive; in some
remedial actions, the contamination  may be  fixed in  place  and
capped, if the hazardous  material can be rendered stable and in-
nocuous.

Contaminated Soil

   It is assumed that highly contaminated soil is excavated  and
removed to a disposal site. Highly contaminated soils are those that
include highly toxic  or carcinogenic substances  such  as PCBs,
volatile organics or pesticides. If the costs for excavation, transport
and disposal are greater than on-site burial in an approved disposal
area, then on-site burial costs are used.

Buildings

   Some sites  have buildings that must be  decontaminated  and
removed. Disposal would either be by on-site burial, removal to an
acceptable waste dump or, if sufficient steel were involved, sale for
scrap. Cost is a function of building floor space and  composition
(i.e., brick, concrete, steel).

Leachate Collection Systems

   At some landfills  and waste dumps, surface and underground
waters  flow through contaminated soils and  materials and carry
dissolved hazardous  materials away from the contaminated area.
                                                        Drainage systems can be constructed around the contaminated area
                                                        to collect the solutions, and a small treatment plant can be installed
                                                        to collect the contaminated liquids for removal from the site or for
                                                        treatment prior to discharge. There are many types of drainage
                                                        systems  and costs will vary from  depth below  the  surface and
                                                        length. By correlating cost data from several installed leachate col-
                                                        lections  systems, the  authors have developed a cost function that
                                                        varies according to the length of  the collection system and  the
                                                        volume of leachate treated.

                                                        Diversion Trenches

                                                          At some sites, surface water that tends to flow into contaminated
                                                        areas (such as landfills or dump areas) can be diverted around the
                                                        area by trenches. The design and length of trenches, dikes or berms
                                                        is site-dependent. The authors have  used some general correlations
                                                        of costs  to develop a  function that  varies by length of the trench.

                                                        Capping

                                                          Impermeable  barriers are  installed above many contaminated
                                                        areas  to  preclude the  flow of  precipitation  through the con-
                                                        taminated area. The authors have considered three general types of
                                                        caps:
                                                        •Clay covered with topsoil and vegetation—general application
                                                        •Clay plus synthetic  liner covered  with topsoil and  vegetation—
                                                         volatile organics and highly soluble chemicals as contaminants
                                                        •Asphalt—roads and  parking areas
                                                          Operating and maintenance costs are assumed at 5% of the in-
                                                        itial capital  costs.  The costs vary  according to the surface area of
                                                        the cap required and were derived  from feasibility studies, engineer-
                                                        ing handbooks and vendor quotation  data.

                                                        Slurry Walls

                                                         Underground  water contamination may be contained in-place by
                                                        installation of a low permeability  wall around the  perimeter of the
                                                        contaminated area. Usually  the wall will extend from the surface
                                                        down to the bedrock. The principal cost parameters are the length
                                                        and depth of the wall. Unit costs appear  to increase as depth  in-
                                                        creases beyond 30 ft.

                                                        Fencing

                                                         All hazardous waste sites, except  those  that  would normally be
                                                        fenced, are assumed to be enclosed by a cyclone fence. Appropriate
                                                        gates and signs are included in the costs.

                                                        Groundwater Treatment

                                                          At some sites, a system of wells will be installed. Contaminated
                                                        water will be pumped to the surface where the hazardous materials
                                                        will be removed and  collected and clean water reinjected into the
                                                        ground.  The number  of wells and the amount of water that can be
                                                        pumped  and processed depend  on the geology of the site and the
                                                        geochemistry, permeability and porosity of the soil. Further, it is
                                                        possible  to vary  the total rate of pumping and the  number of years
                                                        of operation.
                                                          Information on pumping of groundwater is available for only a
                                                        few sites, but a range  of 1,000  to 18,000 gal/day/acre at con-
                                                        taminated areas  has been seen in site design studies. The capital and
                                                        annual operating costs are calculated  based on perceptions of the
                                                        site hydrology (and hence an estimate of solution flow) and the
                                                        types of contamination in the  groundwater which determine the
                                                        treatment process.

                                                        Wells

                                                          Many  of  the Superfund sites include municipal wells that have
                                                        contaminated water.  At numerous other hazardous waste sites,
                                                        public or private drinking water sources are threatened by migra-
                                                        tion of contaminants  in groundwater.  Remedial action in many in-
                                                        stances requires either:  (1) development of alternative water sup-
                                                        plies or (2) treatment of water sources. The cost elements associated
                                                        with these activities may include:
332
COST OF CLEANUP

-------
               NUMBER Of SITES
               CAPITAL COSTS
                     of dollxi)
                                               3-4   4-6
                                                               6-7   7-a
                                                                              9-10  10-11  11-15 1S-20 20-30 30-31
               NUMBER/PERCENT
               OF SAMPLE       13/16% 14/17% 10/12%  5/6% 10/12%  4/6%   6/7% 7/9%   1/1%  1/1%  4/6%  4/6%   2/2%   0/0%  1/1%
                                                             Figure 1
                                      Frequency Distribution of Capital Costs of Remedial Actions
                                                        for Estimated Sites
•A new house well
•The relocation of municipal wells to an uncontaminated area
•The connection of houses to existing municipal water supplies
•The treatment of municipal water

RESULTS

  During visits to USEPA regional offices in Regions 1, 2 and 3 in
late 1983, the authors collected information concerning 88 hazar-
dous waste sites. Of these 88 sites, 82 were amenable to costing by
the system; the  other  six were extraordinary sites such as Love
Canal that had problems that could  not be handled accurately by
the system. Of sites amenable to costing, seven were in Region 1, 38
in Region 2 and  37  in Region 3.
  Regional and state costs of remedial actions at the sites estimated
by the system are given in Table 1. The average capital cost per site
figures exhibit a significant variation between states and regions.
This analysis supports the assumption that it would be difficult to
find a  small sample of sites that would be representative of the
universe.  It also indicates, not surprisingly,  that  there may be
systematic differences in the types of problems experienced at sites
in different regions of the country. For example, for the three states
with the greatest  representation in our sample, average capital costs
per site vary from $8.1 million in New Jersey to $4.9 million in New
York and $3.0 million in Pennsylvania.
  Another aspect of the variability  among costs at NPL sites is
shown in the histogram in Figure 1 which groups sites according to
remedial capital costs. The histogram demonstrates the large range
of remedial costs and suggests that the distribution of costs has its
greatest concentration  at low levels  and declines steadily,  as op-
posed to exhibiting  the more typical  bell  shaped curve of the nor-
mal distribution. Thus the average cost of $5.0 million for the sites
in this sample is the result of averaging a few sites with relatively
high costs with a larger number of sites with relatively low costs.

TOTAL COSTS OF REMEDIAL
ACTIONS AT ALL NPL SITES
  One  of the major  objectives of  this  work was to develop a
methodology that could be used expeditiously and inexpensively to
derive estimates for remedial costs at nearly all sites on the NPL. By
so doing, the need to develop total cost estimates on the basis of ex-
trapolation from relatively few sites could be avoided. In the conti-
nuing effort, a significant number of additional sites will be added
to the database. However, some form of extrapolation was re-
quired to derive a total cost estimated based on the initial sample of
82 sites.
  The extrapolation procedure that was used addressed the two
potential sources of bias in the sample. One potential problem was
that the sample only includes sites that are in some sense "typical"
hazardous waste sites and excludes "extraordinary" sites such as
Love Canal that are not amendable to costing by the system and
where remedial costs are generally much higher. The other problem
                            Table 1
                Cost Results by Region and by State
Region/State
REGION 1
Massachusetts
New Hampshire
Rhode Island
Vermont
Region 1 Average
REGION 2
New Jersey
New York
Region 2 Average
REGION 3
Delaware
Maryland
Pennsylvania
Virginia
West Virginia
Region 3 Average
Sample Average
Number of
Sites

4

3

7

18
19
37

9
3
19
4
3
38
82
Capital Costs
per Site
($ Million)

8.4

4.8

6.9

8.1
4.9
6.5

5.6
2.3
3.0
2.9
1.8
3.1
5.0
Annual O&M
Costs per Site
($ Million)

0.2

0.4

0.3

0.4
0.4
0.4

0.3
0.2
0.2
0.1
0.1
7.0/0.2
22.9/0.3
                                                                                                COST OF CLEANUP
                                                         333

-------
                           Table 2
                    Calculation of Total Cost

Type
I.
f|
S.

4.

5.

6.
7.
8.
9.
10.
11.
12.

13.
14.


of Site
Landfill
Wells
Industrial Dumps
snd Treatment
Chemical Plants/
Refineries
Manufacturing
Plants
Water Bodies
Pure Lagoons
Military Sites
City Contamination
Radioactive Sites
Mines and Tailings
Housing Areas/
Farma
Presque Isle
Extraordinary
Sltea
Number
In Sample
30
7

25

10

4
--
4
1
--
—
1

—
—

—
Number
on NPL
116
49

136

91

95
11
22
3
6
5
8

1
1

(19)
Capital Cost
($ MUllona)
891
233

579

599

333
98
56
4
46
60
71

3
0

950
CUM Coat
(t Millions)
66
24

18

24

19
4
2
0
1
1
3

0
0

29
TOTAL COSTS
                                           3,923
                                                        193
was that the mixture of sites in the sample may not have been
representative of the total mix of sites in the NPL (e.g., too many
landfills, too few manufacturing plants).
  To address the first problem, the authors made additional provi-
sion in the total cost estimates to account for the number of "ex-
traordinary" sites that were anticipated to be on the NPL.
  To address the second problem, each site on the NPL was assign-
ed to one of 13 categories intended to include a number of sites that
would be characterized by problems of similar  origin and nature.
The majority of sites fell into one of the following classifications:
landfills, wells, industrial dumps and treatment  facilities, chemical
plants/refineries,  manufacturing  plants,  lagoons  and water
bodies.* Results from the 82 sites were used to develop estimates of
average costs  for "typical" sites in each  category. These figures
were then used to develop estimates of total remedial costs for each
category and ultimately for the entire NPL.
  The   resulting  cost estimates are  found   in Table  2.  The
preliminary estimates of  capital and first-year  operating  and
maintenance costs associated with remedial actions at all NPL sites
are $3.9 billion and $193 million respectively, or $7.2 million and
$0.4 million per site. However, these results are preliminary and
are, in fact, likely to change as the research progresses. Since the
authors' initial work, the USEPA policy with regard to acceptable
remedial action at sites has moved in a direction that requires more
groundwater treatment and more removal or containment of con-
taminated soil. This emphasis will necessitate revisions in the deci-

*There are many possible groupings for hazardous waste sites and judgment was used in developing
groupings that were believed to be adequate for the purposes of this preliminary extrapolation. Judg-
ment was also used in assigning sites to categories in cases where distinctions were not obvious.
                                                         sion rules used in this sytem that will probably cause an increase in
                                                         these estimates of remedial costs per site.

                                                         APPLICATIONS OF THE METHODOLOGY
                                                           Prior to the commencement of corrective action at any hazar-
                                                         dous waste site,  it is necessary to conduct a detailed analysis of the
                                                         site problems and to evaluate alternative approaches to cleanup. It
                                                         is  mandatory to conduct remedial investigations and  feasibility
                                                         studies prior to  conducting a detailed design of a selected plan to
                                                         correct  site problems.  The sequence of studies  requires a  long
                                                         period of time (i.e., from several months to more than a year) and
                                                         is very expensive (hundreds of thousands of dollars) for each site.
                                                         While it is necessary to follow this approach prior to commitment
                                                         of millions of dollars of construction funds for cleanup at each site,
                                                         there are situations that warrant a more rapid and less costly pro-
                                                         cess to project the cost of cleanup and assess the alternatives.
                                                           This project was initiated to help management and planning per-
                                                         sonnel at the  USEPA cope with the total cost of the Superfund
                                                         Program when a very small fraction of the sites had been complete-
                                                         ly analyzed. This system, with appropriate update, will continue to
                                                         be used as a tool to assess the cost of the Superfund program.  Fur-
                                                         thermore, as government policy changes, this methodology  pro-
                                                         vides a rapid and relatively low cost means to assess the cost impact
                                                         of new policies on Superfund.
                                                           Apart  from use by the  USEPA  for its management and budget
                                                         analysis, the authors believe that this program has value in several
                                                         other areas.  One application is in a preliminary assessment of newly
                                                         discovered or  newly listed hazardous waste sites.
                                                           Using very preliminary  data about the nature of the contamina-
                                                         tion at a site,  it  is possible to estimate quickly the cost of cleanup
                                                         using data  presented here. While this methodology  is not  a
                                                         substitute for detailed engineering feasibility studies and designs, it
                                                         can be used  by industrial firms and responsible panics to assess the
                                                         likely ranges of the  costs of a site cleanup. In addition, this system
                                                         can be used  to rank order the problem areas at each site in terms of
                                                         the approximate cost of the solution.  This ranking would enable
                                                         subsequent studies to focus on the problems that have the greatest
                                                         impact on and contribute  the greatest uncertainty to total remedial
                                                         costs at a site.
                                                           Another application of this methodology would be to assess the
                                                         needs for further research that could significantly reduce the future
                                                         costs of Superfund and for parties responsible for contaminated
                                                         property.  The preliminary results  of this study indicate that the
                                                         most expensive  corrective actions that involve the greatest costs
                                                         across all NPL  sites are:  (1) capping, (2) leachate collection  and
                                                         treatment systems, (3) treatment of underground and surface water
                                                         and (4) slurry walls. Some corrective actions, however, do not have
                                                         established,  engineered solutions  and fundamental  engineering
                                                         design  data are  not available.  Identifying the remedial activities
                                                         that involve the greatest cost expenditures may help target research
                                                         to the areas  that  will have the greatest beneficial impact on the total
                                                         cost and effectiveness of Superfund actions.
334
COST OF CLEANUP

-------
               COST ANALYSIS  FOR REMEDIAL ACTIONS
                                     UNDER SUPERFUND

                                              BRUCE CLEMENS
                                Office of Emergency and Remedial Response
                                   U.S. Environmental Protection Agency
                                               Washington, D.C.
                                          EDWARD YANG, Ph.D.
                                         Environmental Law Institute
                                               Washington, D.C.
                                            BRIAN J. BURGHER
                                                 JRB  Associates
                                               McLean, Virginia
INTRODUCTION

  In  response to risks  posed by uncontrolled  hazardous waste
sites  to public health and environment, Congress  enacted the
CERCLA in December, 1980. The Act authorized the government
to establish the  Hazardous Substance Response Trust Fund (the
Fund) to finance the necessary responses to any release or threats of
release from hazardous waste sites. While the primary goal of the
Act is clearly to protect public health and environment from release
of hazardous substances, it also requires remedial responses to be
"cost-effective" and within the financial limits of the Fund. Cost-
effectiveness considerations must be incorporated into the selection
of appropriate remedial actions with the goal of ensuring the great-
est improvement in protection of public health  and  the environ-
ment for the least cost.
  Cost considerations are a key component in various steps of the
remedial action alternative selection process:
•Screening of alternative remedial measures
•Detailed analysis of the alternatives
•Selection of the alternatives
•Balancing the cost of the selected measures against the availability
 of money in the Fund
  In  addition, the potential for cost recovery from potentially re-
sponsible parties requires a thorough cost documentation.
  In  this paper, the authors  summarize the results  of an effort
undertaken by the USEPA to develop costing information from
which a consistent cost-effectiveness program could be developed.

NEED FOR COST INFORMATION
  The USEPA, responding to the requirements  of the CERCLA,
reviewed existing information regarding available cost data which
could provide: (1)  a basis for future costing and (2) a consistent
approach such that remedial actions could be compared  during a
cost-effectiveness analysis. Cost  information was available from
case studies of remedial actions, direct field construction activities
and standard references. Similarly, cost analysis procedures were
available from other engineering programs.  Unfortunately, the
available information which could be applied directly to uncon-
trolled hazardous waste sites was not adequate for consistent cost
estimations.
  To supplement available information, the USEPA undertook
an effort to:
•Collect relevant, consistent and reliable cost data that could be
 used for cost estimates  for remedial action  at different levels of
 the selection process
•Develop  specific procedures for cost estimating and economic
 analysis required for remedial action planning
  Through these two activities, the USEPA can provide necessary
guidance and baseline cost information for USEPA Regional Pro-
ject Managers, contractors performing  remedial investigation/
feasibility studies,  private  parties undertaking hazardous waste
cleanup activities and State and local remedial action personnel.
Further, concise economic analysis will demonstrate to the public,
Congress  and  other interested  parties  that the USEPA  and
CERCLA are adequately addressing the problems of uncontrolled
hazardous waste cleanup.

COST ANALYSIS TOOLS
  The USEPA's Municipal Environmental Research Laboratory
and Office of Emergency and Remedial Response,  with input from
other USEPA offices, regional offices, environmental organiza-
tions, State and local institutions and private parties, have devel-
oped resource documents to fill the gaps in guidance  for costing
analysis for uncontrolled hazardous waste-site responses. Specif-
ically, the USEPA cost guidance information is  provided in the
following draft documents:
•Guidance for Feasibility Studies Under CERCLA
•Procedures for Cost Analysis for Remedial Action Under Super-
 fund
•Compendium of Cost of Remedial Technology at  Hazardous
 Waste Sites
  These documents are undergoing final review and will be issued
shortly for use by the remedial response community in the develop-
ment of costs for remedial  action planning. The three documents
are examples of three levels of USEPA guidance on CERCLA ac-
tions. The Guidance for Feasibility Studies Under CERCLA is an
example of a "what to" manual.
  Procedures for Cost Analysis for Remedial Action Under Super-
fund is a "how to" manual designed to define the process set forth
in the "what do" manual (Guidance for Feasibility Studies).
  Lastly, the Compendium of Cost of Remedial Technology is an
example of sour information required to follow the  process set
forth in the "how to" manual (Procedures for Cost Analysis for
Remedial Action Under CERCLA). The Office of Emergency and
Remedial Response is preparing a comprehensive annotated list of
all such guidance documents. A discussion of the three costing doc-
uments follows.

Guidance for Feasibility Studies Under CERCLA

  A standard process has been established by the USEPA for eval-
uating remedial action alternatives in terms of cost and  effective-
ness. This process is set forth in the Guidance for Feasibility
Studies Under CERCLA. The Guidance  provides a structure for
                                                                                        COST OF CLEANUP
                                                     335

-------
Regional Project Managers, potentially responsible parties and
others who have responsibility to prepare documentation support-
ing remedial actions performed under CERCLA and the National
Contingency  Plan. As part of the process, a recommended ap-
proach to development and analysis of costs  of  remedial action
alternatives is given.
  The Guidance presents steps  which  should  be  followed  to
develop costs as input to the cost-effectiveness analysis. That is, the
Guidance presents "what to do" rather than "how to" procedures.
Specific procedures are presented in the Remedial Action Costing
Procedures Manual.
Costing Procedures Manual
  The Procedures for  Cost Analysis for Remedial Action Under
CERCLA presents specific procedures for various phases of the re-
medial action planning process. Procedures are provided  to assist
in:
•Preparation of an initial  assessment of remedial action alterna-
  tives to establish a general cost  for the remedial  investigation/
  feasibility study process and initial remedial measures
•Screening of remedial action alternatives during the feasibility
  study to eliminate those alternatives for which the  costs  are sub-
  stantially greater than other alternatives and yet do not provide a
  commensurate public health or environmental benefit
•Preparation of detailed cost estimates for feasibility studies to aid
  in selecting a remedial action alternative
   The Costing Procedures Manual presents procedures and pro-
vides worksheets to accomplish the cost analysis objectives of the
above phases. The guidance presented has been developed for gen-
eralized conditions at uncontrolled hazardous waste disposal sites.
   The Manual defines  procedures  for estimating costs for sites on
the NPL prior to initiation of the remedial  investigation.  At this
stage, an assessment formerly known as the Remedial Action  Mas-
ter Plan (RAMP),  is completed.  It summarizes  existing  site in-
formation, addresses the types of remedial activities required at the
site, addresses community relations concerns at the site and esti-
mates budget and schedule requirements for subsequent remedial
response activities. The USEPA uses this report to plan future site
response actions and to provide general direction to the future ac-
tivities associated with the  remedial investigation and feasibility
study (RI/FS) process.
   The Manual defines the steps of the site response assessment pro-
cess and offers  guidelines for costing during each of the steps.
With this data, the reader can:
•Construct a site outline and identify  areas  which require esti-
  mating
•Assign order-of-magnitude costs to the applicable sections of the
  site response assessment report format
•Determine total order-of-magnitude costs  for site Rl/FS activ-
  ities and remedial alternatives available at this stage prior to the
  RI/FS
   The Manual concentrates on costing  procedures as part of the
Feasibility Study effort.
   Two sets of cost estimates are generated within the overall  Feas-
ibility Study  process. Initially, order-of-magnitude costs  are gen-
erated to screen out disproportionately expensive alternatives. Sub-
sequently, feasibility costs are developed and used to determine the
most cost-effective alternative. While the basic  procedures for gen-
erating  these cost estimates are essentially identical, guidance  is
provided to achieve a greater level of accuracy for the feasibility
costs through the use of more extensive data sources and a more de-
tailed preliminary design based on information available from the
remedial investigation.
   Briefly, screening cost estimates are generated during the alterna-
tive development and screening process. These estimates  are used to
eliminate those  alternatives whose costs are significantly greater
than  competing alternatives yet do not provide commensurate en-
vironmental and public health benefits. Users of the Manual are
directed to the Remedial Action Cost Compendium to  estimate
screening costs along with other sources referenced in  that docu-
                                                        ment. The accuracy of the costs should be in the + 100% tol - 50%
                                                        range.
                                                          Following initial screening, a manageable number of remedial ac-
                                                        tion  alternatives should remain  for  the feasibility cost analysis.
                                                        Cost estimates for feasibility cost  analysis are intended to provide a
                                                        measure of the total  resource costs over time associated with any
                                                        given remedial alternative.
                                                        Cost Compendium
                                                          The  Compendium of Cost  of Remedial  Technology at Haz-
                                                        ardous Waste Sites summarizes existing cost information on typical
                                                        components of remedial actions. Actual  expenditures  and esti-
                                                        mated costs taken from a number of sources have been assembled
                                                        into  this one  data base. The  immediate  use of  this centralized
                                                        source of cost information is to provide consistency in various
                                                        site-specific costing tasks such as remedial alternative costing called
                                                        for in the Guidance  for Feasibility Studies Under CERCLA and
                                                        budgeting for immediate and planned removals.
                                                          The Compendium should be viewed as the first installment of an
                                                        ongoing data base. It will be updated periodically as more cost in-
                                                        formation becomes  available from completed  Superfund  re-
                                                        sponses. Cost data in the Compendium are organized according to
                                                        related technologies such as "Groundwater controls." The costs
                                                        given are  for technologies most  commonly used  at uncontrolled
                                                        hazardous waste sites, although some rarely used technologies are
                                                        given based on engineering estimates. Typically, the number of es-
                                                        timates and  the depth of background information provided  are
                                                        proportional to the frequency of  use of the technology. The Com-
                                                        pendium is a key support to the Costing Procedures Manual. The
                                                        two are constructed to allow detailed interface on every stage of the
                                                        cost estimating.
                                                          The  Cost Compendium contains two  notable features  that
                                                        should be mentioned. First, the Compendium stresses the compari-
                                                        son between estimated cost and actual expenditure from past ac-
                                                        tions. The primary reason  for this is that most of the  engineer-
                                                        ing estimates are not  "field tested" and are based  on normal con-
                                                        struction activities, unrelated to hazardous  waste cleanups. Unfor-
                                                        tunately, there are very few documented actual expenditures  for
                                                        hazardous waste  cleanups. As more remedial actions are com-
                                                        pleted, the data base will become more extensive. Also,  although
                                                        actual expenditures are  in  general more reliable  than estimates,
                                                        there is a great deal of variability in the existing data.
                                                          Second, the Cost Compendium documents the critical factors
                                                        that  often affect  cost estimates  for each  technology.  A  typical
                                                        table of actual cost expenditures from the Cost Compendium is
                                                        shown in Table 1.  Material in this table describes the major factors
                                                        that affected the cost of each of  the slurry walls (including depth,
                                                        length, thickness,  type of material,  etc.) listed in the table. The
                                                        USEPA is in the process of computerizing certain data in the Cost
                                                        Compendium. When a richer data base is available, the USEPA
                                                        will develop regression equations to streamline cost estimating.

                                                        PROCEDURES FOR COST ESTIMATING
                                                          The development of cost estimates for remedial action alterna-
                                                        tives  involves the following steps:
                                                        •Estimation of Costs—estimate capital and annual operating costs
                                                        •Present Worth Analysis—using  estimated costs, calculate  annual
                                                         costs and present worth for each  remedial action alternative
                                                        •Sensitivity  Analysis—the sensitivity of cost estimates to changes
                                                         in various parameters
                                                          Feasibility analysis costs are typically derived from a number of
                                                        sources, including vendor estimates, and should be accurate within
                                                        a range of + 50 to  - 30%.

                                                        Estimation of Costs

                                                          In this section,  procedures  for estimating capital and  annual
                                                        operating costs for remedial  action  alternatives are given. Sunk
                                                        costs are not to be included in this evaluation. Sunk costs include
                                                        investments or commitments made prior to or concurrent with re-
 336
COST OF CLEANUP

-------
                            Table 1
                    Slurry Wall Expenditures
                         (1982 Dollars)
DATA SOURCE
US EPA
EL1/JRB
1981
Pennsylvania
US EPA
JRB/ELI
1979
Colorado
US EPA
JRB
Florida A
(Date unknown)
US EPA
CH2 M Hill
1982
New Hampshire
LENGTH & DEPTH
648 feet
X
17 feet
1 ,500 feet
X
20 feet
2,290 feet
X
30 feet

3,500 feet
X
60 reet
THICKNESS
1 foot
30 inches
30 inches

3 feet
MATERIAL
cement -
bentoni te
85* soil-
ben ton i te ;
15* cement
soil-
bentonite

SOll-
bentoni te
UNIT COST
531.96/sq.ft.
$8.33/sq.ft.
$5.8B/sq.ft.

S5.64/sq.ft.
 medial action planning. The various cost components that should
 be considered and sources for cost data are identified, and work-
 sheets are provided to assist the user in organizing and presenting
 the cost data for each alternative.
   Federal construction programs  have traditionally distinguished
 between capital costs and operation and maintenance (O&M) costs.
 Federal participation in public  works projects  such as  highways
 and wastewater  treatment plants has been limited to construction
 involving the funding of a major share of projet  capital costs.
 Following construction, costs for  operation and maintenance are
 the responsibility of the State or local government.  However, the
 distinction between  the  construction and operation phases of a
 Superfund cleanup response is not as easily made. The completion
 of construction  will not achieve public health  or environmental
 protection in many  instances.  Such protection may be afforded
 only after operation of the remedial technology  for a period of
 time.
   While the distinction between  the remedial  action and O&M
•phases of a  cleanup  is important for determining fund eligibility,
 it should not be a factor in feasibility cost analysis. The Manual
 provides guidance in development of  comparative rife-cycle cost
 information for the  remedial action alternatives under considera-
 tion for use in  the  alternative  selection process. These alterna-
 tives include both the remedial action and O&M phases. Thus, for
 the purposes of  feasibility costing, the user is  directed to observe
 the  conventional distinctions between capital  and  O&M costs,
 where capital and initial construction costs are analogous.
   Capital costs are those expenditures required  to initiate  and in-
 stall a remedial action. They are exclusive of costs required to main-
tain  or operate the action throughout its lifetime and include only
those expenditures that are initially incurred  to develop and in-
corporate a  remedial action (e.g., installation of a  cap  or slurry
trench) and  major capital expenditures anticipated in future years
(e.g., replacement  of a cap or slurry trench).  This differentiation
between capital  costs and  operation and maintenance costs does
not necessarily reflect a determination as to the "Fund"  eligibility
of the costs.
  Direct capital costs  include  equipment, labor  and  materials
necessary  for installation  or construction of  remedial actions.
These include costs for:
•Remedial action construction
•Component equipment
•Land and site development
•Buildings and services
•Relocation of affected population where appropriate
  Indirect capital costs consist of engineering, financial, supervis-
ion and  other services necessary to carry out a  remedial action.
They are not incurred as part of actual remedial actions but are
ancillary to  direct or construction costs. Indirect capital costs in-
clude costs for design and engineering and contingency allowances.
  Operation and maintenance costs are those post-construction/in-
stallation costs necessary to ensure continued effectiveness of a re-
medial action.
  The post-construction/installation activities necessary to ensure
continued effectiveness  of a remedial action may  involve the
following cost components:
•Operating Labor—Includes all wages, salaries, training, overhead
 and fringe benefits associated with the labor needed for post-con-
 struction operations. The user should identify the labor require-
 ments by skill categories for each remedial action alternative.
•Maintenance Materials  and Labor—Include the costs for labor,
 parts and other  materials required to perform  routine  mainten-
 ance of facilities and equipment associated with a remedial action
 alternative.
•Auxiliary Materials and Energy—Include such items as chemicals
 and electricity needed for plant operations, water and sewer serv-
 ice and fuel costs.
•Monitoring Activities—Include costs of sampling, analysis, main-
 tenance of wells and preparation of reports.
•Purchased Services—Include such items as sampling costs, labor-
 atory fees and other professional services for which the need can
 be predicted.
•Disposal—Includes transportation  and disposal of any waste
 materials, such as treatment  plant residues generated during the
 course of a remedial action
•Administrative Costs—Include all costs associated with adminis-
 tration  of  remedial  action operation and  maintenance not in-
 cluded under other categories such as labor overhead.
•Insurance,  Taxes and License—Include  such  items  as: liability
 and sudden and accidental insurance; real estate taxes on  pur-
 chased  land or  right-of-way (for  non-governmental  projects);
 licensing fees for certain technologies; and permit renewal and
 reporting costs.
•Maintenance Reserve and Contingency Costs—Represent annual
 payments into escrow funds to cover anticipated replacement or
 rebuilding of equipment and any large unanticipated O&M costs,
 respectively (for private lease actions).
•Other Costs—Include all other items which do not fit into any of
 the  above categories.
Present Worth Analysis

  This section contains guidance on recommended procedures for
evaluating costs over the planned lives of remedial action alterna-
tives. Present worth analysis provides a method of evaluating and
comparing costs occurring over different time periods by discount-
ing all future expenditures to the present year.
  Present worth analysis is the recommended method of evaluating
expenditures occurring over different time periods. The costs for
different remedial action alternatives can be compared on the basis
of a single figure for each alternative by  discounting all costs to
a common base year. This single  figure—the present worth  or
value of  a project—represents the amount of money which, if in-
vested in the initial year of the remedial action  and disbursed as
needed, would be sufficient to cover all the costs associated with a
remedial action. Worksheets are provided in the Manual to ease
calculation of present worth for cost-effective analysis.
  Cost analysis for Superfund actions should  follow current guide-
lines of the Office of Management and Budget (OMB). OMB Cir-
cular No. A-94 (6) specifies that costs in future years should not be
escalated to  account for general price inflation, except where there
is a reasonable basis for predicting differences in  the relative  esca-
lation of costs (or benefits) associated with the project. Otherwise,
                                                                                                 COST OF CLEANUP
                                                          337

-------
                           Table 2
       Worksheet 7: Summary of Sensitivity Analysis (Example)
       Fiut coin u i.
       f« /*••;
                n.ei?
                     11,
              , «>•/
                                    Jvf
                                        '< "Y
                                                  \'t^f
                                               fin
                                                     ts*.
                                                1  I
the analyst should use constant (i.e., base period) dollars. Given
the difficulty in forecasting relative price changes over an extended
period, such forecasts should not be included as part of F.S. Cost
Analysis, except perhaps as part of the sensitivity analysis. OMB
currently specifies a discount rate of 10%, which represents "the
average rate of return on private investment before taxes and after
inflation".
  The period over which a  remedial action requires maintenance
and/or operation (period of performance) is also an important fac-
tor in present worth analysis. Remedial action alternatives requir-
ing perpetual care should not be costed beyond 30 yr for the pur-
pose of feasibility analysis. Present worth  of costs beyond this
period becomes negligible and has little impact on the total present
worth of an alternative.

Sensitivity Analysis

  Procedures are provided for evaluating the sensitivity of cost fig-
ures to changes in assumptions. Table 2 is an example of a work-
sheet provided to document the sensitivity  analysis.
  The  following factors are  recommended for consideration in
conducting sensitivity analysis:
•Effective Life of Remedial Action. If the remedial action alterna-
 tive relies on a new technology or a technology that has not been
 tested over a full demonstration, the analysis should consider the
 possibility that  all or a portion of the technology may need to be
 replaced during the life of the remedial action. In estimating re-
 placement cost, use base period dollars;  do not adjust for infla-
 tion.
•O&M Costs.  O&M costs, if required, are likely to represent a
 substantial portion of total project cost because they may be re-
 peated each year for as long  as 30 yr. The major components of
 O&M cost should,  therefore, be considered  for examination in
 the sensitivity analysis.
•Duration of Cleanup. The duration of cleanup, or period of per-
 formance, is often a key variable (e.g., in actions that require the
 operation of treatment systems for a period of time based on mon-
 itoring results). Various assumptions about the length of period
 of performance may be suitable candidates for analysis.
•Uncertainty Regarding Site Conditions. Even after the conclusion
 of  a remedial investigation,  significant  uncertainties may exist
 regarding  the extent of cleanup necessitated by site conditions.
 Examples are the volume of groundwater to be treated, the num-
 ber of drums to be excavated, the type of materials present and
 the treatment disposal options to be used. Various assumptions
 regarding such parameters may need to be examined.
•Inflation. Inflation should not generally be examined under OMB
 costing guidelines, but it may  be considered under two conditions:
 first, if there is good reason  to believe that the future prices of
 materials or services  required by a  remedial alternative will in-
 crease at a significantly faster or slower rate than the general level
 of prices in the economy; or  second, if the inflation rate for the
 area in  which the site is located can be expected to vary signif-
 icantly from the national average.
•Cost of Borrowed Capital. In private or State actions, the cost of
 capital may be a major factor in determining the overall cost of a
 remedial action and may, therefore, be tested in the Sensitivity
 Analysis.
•Discount Rate.  When comparing alternatives that have major
 differences in the mix of O&M  and capital costs, determine
 whether a 5% discount rate changes the relative costs of alterna-
 tives.
338
COST OF CLEANUP

-------
              A REVISED  COST  MANAGEMENT APPROACH
                              FOR SUPERFUND  REMOVALS
                                              JAMES R. JOWETT
                                         Emergency Response Division
                                    U.S. Environmental Protection Agency
                                                Washington, D.C.
                                              ROBERT J.  MASON
                                                  SPER Division
                                               Roy F. Weston, Inc.
                                                Washington, D.C.
                                                In Association with
                  Tetra Tech, Inc.  • Jacobs Engineering Group,  Inc.
                   ICF Incorporated
INTRODUCTION
  The Comprehensive Environmental Response, Compensation
and Liability Act gives Federal On-Scene Coordinators  (OSCs)
authority and a funding base to respond to hazardous substance
emergencies that threaten  public health and/or the environment.
The OSC's  primary responsibility at Superfund-financed emer-
gency responses (removals) is to monitor or perform all contain-
ment, cleanup and disposal activities necessary to protect public
health and the environment. Given that OSCs are responding to
environmental emergencies, their responsibilities take an added de-
gree of importance.
  However, OSCs' responsibilities  go beyond those described
above. Removals involve substantial CERCLA funding; they are
also under  intense  public scrutiny.  The typical  removal costs
$150,000 per week  of operation. Therefore, OSCs must be effec-
tive cost managers.  In addition, CERCLA requires that reimburse-
ment is sought from responsible parties for  all Federal response
costs undertaken to implement the Act. Subsequently, OSCs must
also document the  costs associated with removal actions for pur-
poses of cost recovery.
  The purpose of this paper is to propose a state-of-the-art ap-
proach  to cost management that not only efficiently fulfills the
OSCs' cost management  objectives, but also aids  in improving
technical decision-making. The approach is based on the results of
a study of current cost control procedures. The proposed approach
consists of:
•Defining all costs which must be tracked during a removal action
•Developing mechanisms to track all costs
•Streamlining existing cost management procedures
•Evaluating the use of portable microcomputer systems
OVERVIEW OF REMOVAL COST MANAGEMENT

Goals of Removal Cost Management
  OSCs' cost management responsibilities are based on the tenet
that cost management goals are most effectively achieved at the
removal site. OSCs have three cost management goals at a removal
action. The first goal is to track costs for cost control purposes.
This goal ensures that cleanup resources are used effectively and
efficiently  to avoid unnecessary costs. To accomplish this goal,
OSCs must identify and  coordinate resources available through
government agencies and  commercial contractors to ensure effec-
tive response action, and then track  costs to ensure that resources
are used effectively and efficiently to avoid unnecessary costs.
  The second goal is to estimate  costs for cost projection. This
goal has two elements. The first element is to ensure accurate cal-
culation of anticipated removal action costs prior to the start of a
removal. If this goal is achieved,  OSCs will be able to calculate
accurate removal cost ceilings. The second element is to ensure an
accurate running total of actual expenditures and estimates re-
sources necessary to complete the removal action.  Ongoing cost
projection, if successful, provides advance warning of any need to
increase the removal project cost ceiling. To accomplish this goal,
OSCs must project costs to determine initial removal cost-ceilings
and then provide advance warning of any need to increase the ceil-
ing.
  The third goal of removal cost management is to maintain cost
records for cost recovery. This goal requires OSCs to maintain cost
records that can  serve, when needed, as adequate documentation
of expenditures  for  cost  recovery actions  against responsible
parties. To accomplish  this goal, OSCs must maintain accurate,
legally  defensible cost records to serve as adequate documenta-
tion of expenditures for cost recovery actions against responsible
parties.
Analysis of Existing Cost Management Approach
  The USEPA Emergency Response Division  (ERD) is responsible
for managing the removal  segment of CERCLA's response pro-
gram. To ensure consistent implementation of cost tracking and
cost controls at all removal actions, USEPA-ERD issued the Cost
Control Manual for Superfund Removals in June, 1982, to provide
guidance to OSCs on how to track and control costs at CERCLA
removals.
  In March, 1983, USEPA-ERD  directed the Roy F. Weston
Headquarters Technical Assistance Team (TAT) to evaluate the
effectiveness of the removal cost tracking and  control process since
the distribution of the Cost Control Manual. The assignment in-
cluded  ascertaining  how  all   participants—in USEPA   Head-
quarters, in the Regions and in the field—currently track and con-
trol costs and identifying how cost management procedures may
be improved.
  In conducting the  analysis,  TAT interviewed  USEPA  Head-
quarters personnel in  the Emergency Response Division, Financial
Management Division, Procurement and Contracts Management
Division, Office  of the  Inspector General, Office of Waste Pro-
grams Enforcement and Office of Enforcement Counsel. TAT also
interviewed USEPA Regional ISCs and TATs as well as USCG
Strike Team members. In addition, TAT reviewed all applicable
guidance and policy documents. Finally, TAT reviewed completed
removal action files and conducted removal  site visits to observe
on-going cost management procedures. Based on the results of this
analysis, the USEPA-ERD determined that a more sophisticated
cost management system reflecting an evolving removal program is
needed for the following reasons:'
•New CERCLA cost  pollicies have evolved since the Cost Control
 Manual was introduced
•Adequate numbers of personnel often are not available to assist
 OSCs in carrying out removal cost management assignments
•Cumbersome cost management  administrative tasks take too
 much time to complete, thus detracting personnel from other on-
 scene responsibilities
•Access is often not available  to  data that provide OSCs with
 sufficient technical  information  prior to determining  response
 mitigation techniques
                                                                                        COST OF CLEANUP
                                                     339

-------
                            Table 1
                 Desirable Field Computer Features
      Computer Characteristics
      40 IbTolal Weigh!


      Designed Tor Durability



      >256KRAM


      3.5 in. Diskettes

      8088 or 8086 CPU

      MS-DOS

      80 Character by 16 Line Screen


      300/1200 Baud Modem
      Plain Paper Primer
                   Easily transportable; may be carried onto
                   airplane
                   Must withstand advene Held conditions;
                   dust, smoke, heal and shipment from site to
                   site.
                   Ample amount for largest software applica-
                   tions anticipated.
                   Much more durable than 3.25 in. diskettes;
                   easier to work with, less likely to fail-
                   rasl microprocessor; IBM compatibility re-
                   quired by OIRM
                   A USEPA-compatible disk operating system
                   Adequate screen sire for readability; func-
                   tional for menu-driven programs, though a
                   larger screen Is preferred.

                   Standard telecommunications \peeds.

                   Required for enforcement purposes to pro-
                   vide hardcopy documentation of site costs
                   and ovliviticv
  Subsequently, USEPA-ERD and TAT are developing a compre-
hensive approach to removal cost management that should reflect
all removal cost elements, ease the administrative burden to OSCs
on-site, improve technical decision-making and fulfill the goals of
cost management.

Proposed Cost Management System
  The cost management system incorporates two main functions:2
•Planning prior to actual removal action
•Monitoring on-scene contractors and Federal, State and  local
 agencies providing services once removal action work has begun
  OSCs going to the field should have a management plan ready to
implement.  Preparation of  this  management plan begins well in
advance of any removal action. OSCs must be able to identify and
coordinate the resources  available to  him/her through  response
agencies and commercial contractors and develop cost information
through both  general preplanning and incident-specific  planning.
Planning is  accomplished  well in advance of any removal and in-
cludes the following tasks:
•Becoming  familiar with private contractor abilities  to  clean up,
 transport and treat, store or dispose of hazardous substances
•Identifying available support resources
•Developing a generic safety plan
•Analyzing previous USEPA costs associated with hazardous sub-
 stance response actions
•Preparing incident-specific cost estimates
  Once work  has begun at a removal, on-site cost  management
procedures must be carried out and completed in detail.  OSCs are
responsible for the monitoring of contractors and Federal, State or
local  agencies providing services to ensure that cleanup objec-
tives and control of on-site spending are achieved. Monitoring on-
scene response services is a daily responsibility that  includes the
following tasks:
•Issuing daily work assignments
•Documenting (using  logs and records) personnel,  equipment and
 materials used
•Overseeing cleanup activities
•Reconciling contractor reported costs with OSC records
  The revised cost  management system builds upon  the existing,
albeit incomplete, cost control process. Key improvements  to the
system include mechanisms to ensure tracking of all cost elements
that are potentially recoverable from responsible parties and/or are
included in  the removal project  ceiling. For example, where site-
specific cost elements  cannot be tracked in the field (e.g., USEPA
contract laboratory costs), sufficient on-site information is main-
tained so that these costs can be tracked and accessed  elsewhere
for purposes of cost recovery.
  A cost projection system provides a  comprehensive way to pre-
dict and track  all costs attributed to removal project ceilings. The
cost projection system provides an approach for OSCs to estimate
the major costs of a removal given the proposed removal action
to be taken, length of time and other circumstances. A factor of
15% of the sum of the major costs is added to account for remain-
ing removal costs. Using the same approach, cumulative costs are
maintained on a daily basis.
  Also, emphasis is placed on using available support resources, to
the extent practicable, to oversee cleanup contractor performance.
Oversight responsibilities include observing contractor work (on a
full-time basis whenever necessary), reporting problems to the OSC
and ensuring that contractors adhere to safety protocols. Adequate
cleanup contractor oversight may avoid three significant problems:
cost overruns, inadequate work and jeopardized worker safety.
Computer Applications At Removals

  In order  to  streamline  cost management tasks,  the USEPA-
ERD and  TAT  have been conducting an evaluation  with  the
USEPA  Environmental Response Team (ERT) to determine  the
utility of providing OSCs with portable microcomputers to aid in
fulfilling removal cost management and technical objectives. The
idea to use portable microcomputers at removals evolved because
documenting costs is time consuming; much  data are generated
and duplicated; and accurate records must be maintained. If a cost
management system could  be computerized efficiently, cost  in-
formation could be readily prepared, and  more time would be
available to  perform other field duties including monitoring clean-
up  activities. In addition, the portable microcomputers could be
used to access  and manipulate technical data that would  aid in
making competent technical decisions, thus improving the environ-
mental integrity of USEPA response  actions  and potentially  re-
ducing costs.
  The evaluation results demonstrated sufficiently that the integral
use of portable microcomputers at removals could assist in  fulfill-
ing cost management and scientific goals while easing the workload
of OSCs and support personnel.1 The field-based computers and
peripherals tailored to remove action needs could simplify storage,
computation and dissemination of cost information and other per-
tinent information, broaden and simplify access  to scientific and
cost data bases and improve decision-making capabilities of  OSCs.
As  a result, better cost and  scientific management  could  be
achieved at removals. At removal actions, computers could be used
for:
•Projecting initial and ongoing removal costs
•Automating word processing
•Generating daily cost management records
•Accessing the electronic mail system
•Accessing on-line and  proposed scientific data bases such as
 OHMTADS, removal and remedial case history files and data on
 hazardous  waste  transporters  and  treatment/storage/disposal
 facilities
•Storing, manipulating and retrieving  field-generated scientific
 data
  The USEPA-ERD and USEPA-ERT have evaluated portable
microcomputer features which should be considered when selecting
computers for removal program field use (Table 1).
CONCLUSIONS
  The requirements of an evolving removal program justify the
use of state-of-art approaches to removal cost management. The
USEPA and TAT are in the process of revising removal cost man-
agement procedures. In addition, options to computerize these pro-
cedures are also being evaluated. Upon completion of these tasks,
field personnel will  be trained to  use the new cost management
system.

REFERENCES
1.  Mason, R.J.  and Nielsen,  S.E., Review of the Suptrfund Removal
   Cost Tracking/Cost Control System. Roy F. Weston, Inc., 1983.
2.  Mason, R.J., et al., Superfund Removal Program Cost Management.
   RoyF. Weston, Inc.,1984.
3.  Mason, R.J., et al.. Evaluation of Computer Equipment for Removal
   Cost Management. Roy F. Weston, Inc., 1984.
340
COST OF tLEANUP

-------
     FACTORS INFLUENCING CLEANUP  COSTS DURING A
       SUPERFUND REMOVAL ACTION: A  CASE HISTORY

                                              LT.  W.D.  ELEY
                                              U.S.  Coast Guard
                                            Marine Safety Office
                                             Port Arthur, Texas
                                             LT. T.A. BAXTER
                             National Oceanic &  Atmospheric Administration
                                         University of New Orleans
                                          New Orleans,  Louisiana
CHRONOLOGY

  While horse back riding on February 2, 1984, Gerry Foreman
noticed orangish-green liquid standing in a wide,  shallow road-
side ditch adjacent to his property in a rural area of Orange Coun-
ty, Texas  (Fig.  1). A  Texas  Department  of Water Resources
(TDWR)  investigator, who was called by the Foreman family,
surveyed the ditch and estimated that several thousand gallons of
an unknown water solution were in the ditch for  approximately
one-half mile along State Highway 105. TDWR requested a nearby
DuPont Chemical Plant's assistance in identifying the pollutant.
Chemists from DuPont took samples of the liquid and by the even-
ing had tentatively identified the contaminant as a 5 % solution of
sodium chromate or dichromate.
  The responsible Federal Agency, the USCG Marine Safety Office
in Port Arthur, Texas was notified. The Foreman  Road location
was within the coastal zone of responsibility given to the Coast
Guard under the National Oil and Hazardous Substance Pollu-
tion Contingency Plan.  In addition,  there was a great likelihood
of the spill reaching navigable water since the drainage ditch was a
                        Figure 1
                   Location of the Spill
tributary of Cow Bayou, a navigable waterway  of the United
States.
  The liquid was evaluated as an illegal dump of sodium chromate.
Once the chemical was identified and it was established that the
responsible spiller could not be readily identified, the USCG, acting
under an emergency  removal situation, called in an  industrial
vacuum service to collect the liquid. CERCLA funds were accessed
for emergency response and immediate removal action. Upon close
examination, it appeared that at least one tank truck load (approx-
imately 5000 gal) of concentrated sodium dichromate solution had
been dumped into the ditch some time earlier (Fig. 2). Evidently,
this illegal dump of material had, over a period of several days,
moved down the gentle gradient of the ditch for one half mile. Be-
fore discovery, an unknown portion of the hazardous material had
entered Cow Bayou, a tributary of the Sabine River.
  The initial actions were directed toward preventing further run-
off of contaminated water. Sand dikes were placed at various loca-
tions along the half mile of ditch. Cleanup crews worked contin-
uously,  vacuuming  the standing  contaminated  water which con-
tained dichromate compounds in concentrations from 50,000 to
150,000 mg/1.  Coast  Guard personnel monitored the removal,
supervising the  work and providing waste control manifests  for
disposal of the liquids at Chemical Waste Management's West
Port Arthur, Texas landfill, 25 miles distant.

Water Removal

  By Feb. 4, 20,000 gal of contaminated water had been removed.
A  slow drizzle of rain had  hampered cleanup and increased  the
amount of water requiring disposal. In addition, an elevated water
supply canal bisecting the site contributed a constant influx of
water. Low spots of an adjacent pasture were saturated with water.
It was feared these areas were contaminated or would become con-
taminated with further rain.
  Under these  adverse conditions, enough water was removed by
the next day so that contaminated soil, although saturated, could
be excavated. Vacuuming of contaminated water was stopped. The
low lying pastures were fenced and, at the request of the OSC, state
health officials were called in to take milk samples from cows in an
adjacent pasture. Bulk solid waste trucks were leased by the con-
tractor to haul the soil that would be excavated. From consulta-
tion with the  on-scene NOAA  Scientific Support Coordinator
(SSC) and soil chemists from the Research Planning Institute,  a
long term and  immediate cleanup plan was formulated. (For the
duration of on-site activity,  the SSC continued to  monitor clean-
up progress, consulting with outside experts as necessary.) Using
the results of initial site sampling, it was decided to remove the top
                                                                                      COST OF CLEANUP
                                                    341

-------
18 in. of ditch bottom in the area of highest concentration or the
area immediate to where the  chemical  had  apparently  been
dumped. Once the area of high contamination was removed, soil
could be removed more leisurely from the rest of the ditch.
  By Feb. 6, 60,600 gal of liquid and 110 yd1 of soil had been re-
moved at a  cost of $45,000. The contractor was using a vacuum
truck to suck water just ahead of a large mechanical shovel. The
shovel was removing the top 18 in. of soil and placing it into lined
truck beds  for carriage to the  disposal site. By Feb. 8, 300 ft of
ditch had been cleaned. This area  of excavation constituted that
portion of the ditch where the liquid pool formed after the dump.
  It was apparent from the color of the clay remaining behind after
excavation that further excavation was needed. The reason for the
vertical penetration of the sodium chromate was at first unclear. In-
put  from soil  hydrologists  had indicated that  chromate solution
chromium would migrate vertically through the local  Beaumont
clay overburden at a rate of approximately 2 ft/yr.

The Rain Comes
   On the afternoon of Feb. 9, 4 in. of rain fell in a period of less
than 2 hr. Thunderstorms had been predicted, so large sand dikes
had been built to prevent runoff. But water from the deluge of rain
and runoff  from the extensive area drained by the roadside ditch
broke through the dike in a matter of minutes.
   On Feb.  11,  approximately 3 additional inches of rain fell. Re-
sponse  personnel and contractors could do little more than watch
as millions of gallons of rain water washed the length of the ditch.
On  the three  following  days, the ditch  continued  to drain the
flooded pasture and  woodland  areas above the drainage ditch.
Samples analyzed indicated less than 1 mg/1 chromium in the run-
off.
   Media attention given the apparent illegal dump had  heightened
public interest. Unrelated cattle and other animal deaths in the area
were attributed to the dichromate poisoning. Investigation of the
deaths  was  diverting  manpower from  the cleanup effort. In an
attempt to  dispell the concern of  local residents, the  Center for
Disease Control Environmental Health  Section  was  consulted.
Local  residents were interviewed, and samples of well water and
locally  produced  dairy  products  were analyzed for  chromate.
Collective concern for public  health in the spill area never ma-
terialized. Laboratory tests and statements of health  officials were
widely  publicized in an effort  to reassure the public. In addition,
the  lexicological properties of sodium dichromate were thorough-
ly explored in  daily news releases during the  initial days of the
cleanup.
   On Feb.  15, the ditch was rediked. One sand dike divided the 500
ft of ditch that had been excavated prior to the rains.  The dike sep-
arated several thousand gallons of rain water from the most con-
taminated soils. It was soon apparent from the changes in water
color that chromate compounds were going into solution from the
clay.
   Approximately 35,000 gal of rainwater were contained in exca-
vated areas of the ditch. The  level of chromate of this water in-
creased from 50 to 600 mg/1 in less than 24 hr.

Soil Removal
   After dewatering the 500 ft of excavated ditch on Feb. 17, 1984,
it was discovered that crayfish burrows had allowed concentrated
dichromate  solutions to penetrate approximately 4 ft beneath the
original ditch bottom. Sodium dichromate solution had entered the
burrows and then migrated laterally 6 to 8 in. These cylinders of im-
pregnated clay were apparently responsible for the contamination
of thousands of gallons of rain water.
   On Feb. 18,  an additional 80 yd' of clay were removed. The cray-
fish burrows and the surrounding contaminated clay were removed
by hand. These burrows occurred on an average of one burrow/
4 ft2 of ditch bottom; 250  ft of ditch bottom  was treated in this
manner.
   After excavation, the site was capped with crushed shell to main-
tain the underlying clay in an  alkaline  condition, encouraging
                                                                                  Figure 2
                                                                           Site Plan of (he Spill Area
                                                       stabilization of the less hazardous trivalent chromium. A 1 ft layer
                                                       of clay was packed over the crushed shell to limit surface migra-
                                                       tion and leaking effects. This buffering and capping was intended
                                                       to seal remaining chromates in the soil.
                                                         Capping the most contaminated area of the ditch completed the
                                                       most crucial stage of the cleanup. The site was now stabilized. No
                                                       more rain water would  be contaminated. Ninety percent of the
                                                       sodium dichromate contaminants had been removed. This  site
                                                       stabilization stage took  16 days and required the expenditure of
                                                       $90,000.
                                                         Soil samples taken at regular intervals along the 2500  ft of re-
                                                       maining ditch were found to have from 85 to 300 ppm chromium.
                                                       It was determined that removing the top 10 in. of soil from this sec-
                                                       tion of ditch  would reduce  the chromium contamination to the
                                                       background level of 2 to 3  ppm. Cleaning this portion of the ditch
                                                       would require the removal of 1200 yd1 of soil at a cost of $120,000.
                                                       However the operations could wait for the best weather conditions.
                                                       Excavating 200 ft of ditch bottom per day,  the removal  of all
                                                       known contaminated soil was completed on Feb. 29. The ditch was
                                                       backfilled using spoil supplied by the Texas Highway  Department.
                                                       A small bulldozer was used  to restore the gradient and pack the
                                                       soil.
                                                         Soil borings in the ditch and well water were analyzed  in April
                                                       1984. No samples had chromium levels above background level.
                                                         All site activity was completed on Apr. 20, 1984. The total cost
                                                       of the cleanup,  excluding  the use of USCG  resources,  was
                                                       $218,000. Removed were 105,430 gal of liquid and 1465 yd' of soil
                                                       as hazardous waste.

                                                       HAZARDS OF SODIUM DICHROMATE

                                                         Sodium dichromate is a  hexavalent chromium salt fl^C^O?).
                                                       Hexavalent chromium salts are used extensively in metal  pickling
                                                       and plating operations, in  anodizing aluminum and in the manu-
                                                       facture of paints, dyes, explosives,  ceramics and paper.' In the
                                                       petrochemical  and oil  refining industry  complex  surrounding
                                                       Orange, Texas,  sodium dichromate is often used as an additive in
                                                       cooling towers to inhibit corrosion. It can also be used in drilling
                                                       muds for high temperature  oil exploration.
                                                         Hexavalent  chromium has carcinogenic potential.  Water insol-
                                                       uble (IV) compounds have been assigned a threshold limit value
                                                       (TLV) of 0.05 mg/m' by the ACGIH1. Water soluble hexavalent
                                                       chromium compounds have likewise been assigned a TLV of 0.05
                                                       mg/m' although their carcinogenic potential has not been demon-
                                                       strated. Inhalation of dust  will cause respiratory irritation and, in
                                                       severe cases, nasal septal perforation.  Ingestion causes vomiting,
                                                       diarrhea and,  more rarely,  stomach and kidney complications. Re-
                                                       peated skin contact will cause dermatitis.'
                                                         Sodium dichromate is toxic to bluegill fish with a median toler-
                                                       ance limit (TLm) of 145  mg/l over a 24 hr period in fresh water.'
                                                       Crayfish are much hardier. USCO personnel  recovered numerous
                                                       crayfish  at the spill site that had survived in 5 to  10% sodium di-
                                                       chromate solutions for over two weeks with no apparent ill effects.
                                                       One of the crayfish was analyzed and found to have a total chrom-
                                                       ium concentration of 0.86 ppm.
 342
COST OF CLEANUP

-------
  USCG officials  supervising cleanup  operations  on Foreman
Road  dictated that, during the removal of contaminated water,
chemical resistant boots, gloves and coveralls be used by all per-
sonnel.  During soil excavation, when sodium dichromate  dust
could  be generated, approved paniculate filters were an additional
requirement.

FACTORS INFLUENCING CLEANUP COSTS
  It quickly became apparent that the Foreman Road  cleanup
would be expensive. An initial cost ceiling for superfund expendi-
tures was set at $10,000. Within days after initial response the cost
ceiling was boosted to $20,000, then $200,000 and finally $250,000
as the laboratory results revealed the extent of soil contamination.
  The Coast Guard On Scene  Commander (OSC) also  realized
that several internal and external factors, some which could be con-
trolled and some which  could not, would potentially escalate clean-
up costs well beyond the  proscribed $250,000 bench mark. The
range of costs of the various components of the Foreman Road
cleanup are illustrated in Figure 3.
  Although  mitigation  and cleanup of the Orange County, Texas
spill were  straightforward  and uncomplicated, the  range of total
cleanup costs could have varied greatly. Figure 1 shows this dis-
parity. For example, disposal costs  could have been a minimum of
$121,000 if all liquids had been deep well injected and weather con-
ditions had remained favorable. A maximum of $173,000 in dis-
posal charges would have accumulated if the various levels of con-
taminated water had not been segregated or  standing water had
not been allowed to evaporate during fair weather.
  The major  economic considerations  of  the federal  response
action were  selection and use  of the  cleanup contractor, dis-
posal methods, weather conditions, the methods of excavation and
backfill and the sampling  program. Each of these economic con-
siderations will be considered in subsequent sections of this paper.
                                                284.7
                                                  2171
                                                  IBS
             173.5
              1421
              121
                          54.6
      •        n
 e
 3
1.5
    LAB
 ANALYSIS
                        41
                                  27.6
                DISPOSAL
   PRIME        SUB
CONTRACTOR  CONTRACTOR
TOTAL
COSTS
            REPRESENT* WHAT WAS ACTUALLY SPENT


            IHIOH t LOW NUMBERS REPRESENT THE CONCEIVABLE RANOE OP
            EXPENDITURES IP OTHER VARIABLE* THAT AFFECT COST HAD
            BEEN CONSIDERED OR NEQLECTED
                            Figure 3
       Breakdown of Cleanup Costs at the Orange County, Texas
                 Sodium Dichromate Spill in $(000)
  Secondary economic considerations were use of USCG super-
visory personnel, site safety and  health, use of on-scene equip-
ment, quantitive analytical techniques, public concern and media
attention.

SELECTION AND USE OF THE CLEANUP CONTRACTOR
  The Port Arthur Marine Safety Office did  not have a contract
with a firm for hazardous material cleanup and disposal. In the
hours immediately following discovery  of the spill, the  USCG
turned to its federal oil pollution cleanup contractor. This com-
pany, an industrial vacuum service, had been awarded a compet-
itive bid contract to  cleanup oil pollution occurring in navigable
waters of the United States where the source was either unknown
or the spiller refused responsibility for cleanup.
  Given the  requirements  of  the Foreman Road cleanup, this
vacuum  service  worked well.  Removing standing contaminated
water, excavation, soil disposal, backfill  and grading were easily
accomplished by the contractor's personnel and subcontractors.
  Two days into the cleanup, the USCG invited two other com-
panies specializing in hazardous material  removal  to visit the site
and bid on finishing the cleanup. Both firms declined to bid on the
project, realizing that they could do little more than the contractor
already on-site.
  Through previous work in oil pollution cleanup, the MSO had
developed confidence and trust in the abilities of  the contractor.
However, if the  pollutant had  had other  hazardous properties in
addition to moderate toxicity, major obstacles would have had to
be dealt with. The contractor did not have the pumping and con-
tainment systems for corrosive and highly flammable substances.
Personnel did  not have protective equipment or training in haz-
ardous material site safety. If the contaminate had the toxicity of,
for example, arsenic  trichloride or tetraethyl lead, the govern-
ment's oil pollution  contractor could not have been safely em-
ployed.
  Financially,  using the government's cleanup contractor worked
to the MSO's advantage. The oil pollution cleanup Basic   Order-
ing Agreement (BOA) that the vacuum service was under was not
valid for sodium dichromate removal. Another contract had to be
written. But, because the contractor had  been on-site for several
days before the contract could be finalized, the solicitation could
be written "job  specific"  and thus avoid the inventory require-
ments typical of  a contract written to cover response for a variety
of chemical spills. Contractual mandates for existing inventories of
specialized response equipment usually escalate rates for  cleanup.
However, one disadvantage of a  contract written for a cleanup
company already on-site is it will not undergo the competitive bid
process.  Services  were rendered at the going rate and did not in-
clude reductions in equipment and personnel costs that would be
expected from the low bidder.  Moreover, government contracts
normally restrict or eliminate overtime and subcontracting  service
charges. This type of financial  curtailment, of course, could not
be levied on a contractor  already working for the government
under good faith and in the midst of stabilizing an uncontrolled
waste site.
  It would be difficult to determine whether money could have
been saved if the USCG had a contract with a hazardous material
cleanup company prior to the Foreman Road incident.  As men-
tioned,  this  hazardous waste required neither specialized equip-
ment nor protective clothing. An oil pollution cleanup contractor
easily performed  the necessary work. But in the long run, over  a
series of federal hazardous material emergency response actions, it
undoubtedly is in the best interests of the taxpayer to award a com-
petitive bid contract to a cleanup service.

DISPOSAL
  The disposal of chromate contaminated water and soil amounted
to 74% of the total  cost of the Foreman Road cleanup.  Of the
$160,665 spent on disposal, only $34,473 or 16%  of the total bill
was spent on transportation of waste from the spill site to the dis-
                                                                                                COST OPTCLEANUP
                                                                                                                        343

-------
posal site. So approximately $126,000 was expended for disposal.
Since the only hazardous property of the contaminated material
was  its moderate toxicity, the preparation  of the waste for dis-
posal centered around its  physical properties, especially density,
rather than chemical properties like flammability or reactivity.
  Hazardous waste  arrived at Chemical Waste Management's
Port Arthur disposal site as either contaminated water or soil. If the
liquid contained a minimum amount of suspended solids,  it was
deep well injected for 18
-------
 posal options? Does this person have the technical expertise to
 assist you or the cleanup contractor in preparing waste for eco-
 nomic disposal?
•Have you identified a hazardous material cleanup  contractor in
 your area? Do you have a contract with them? Would a compet-
 itive bid contract be appropriate?
During a Cleanup
•Have you invited other cleanup firms to the site to offer a bid?
 Have you at least discussed the cleanup with other firms? Do you
 have alternative methods of cleanup?
•What local industries can help you? Can they provide quantitive
 analysis?
•Are you regularly reconsidering the effectiveness of cleanup and
 disposal methods?
•Are you consulting with the contractor and a waste disposal fa-
 cility regularly?
•Will adverse weather conditions ruin  cleanup  efforts? Can you
 afford to wait for better weather?
•Have you consulted USEPA and other appropriate  governmental
 agencies, not just initially but at regular  intervals  for the entire
 cleanup?
•Is media attention and public concern making you do more work
 than is necessary? Are you talking to the press regularly to  keep
 them updated and  help alleviate unreasonable concerns in the
 community?
DISCLAIMER

  The views expressed in this paper are those of the authors and do
not necessarily reflect the views of the U.S.  Coast Guard or Na-
tional Oceanic and Atmospheric Administration.
REFERENCES

1. McKee, J.E. and Wolf, H.W., Water Quality Criteria. The Resources
  Agency of California, 1963, 163-166.
2. Threshold Limit Values for Chemical Substances in the Work Environ-
  ment for 1983-84, American Conference of Governmental Industrial
  Hygienists.
3. The Chemical  Hazards Response Information System (U.S. Coast
  Guard Commandant's Instruction M16465.12), VOL 2.
                                                                                               COST OF CLEANUP
                                                         345

-------
      STATISTICAL  CONSIDERATIONS  IN  GROUNDWATER
                                              MONITORING

                                               SEONG T. HWANG
                                        Department of  Civil Engineering
                                                Howard  University
                                                Washington,  D.C.
INTRODUCTION

  This paper examines statistical aspects of obtaining groundwater
samples  from monitoring wells  at  hazardous  waste sites and
analysis of the chemical concentration data. Ever since the pro-
mulgation of the requirements by the USEPA  in the 1970s for
monitoring groundwater quality at hazardous waste sites, there has
been uncertainty in how to interpret the monitoring data obtained
and how to ensure high quality data.
  There are several versions of statistical tests proposed for use in
determining  groundwater  contamination.  The  use  of  these
methods, however, has been reported to cause high rates  of false
positives.
  The emphasis  in this paper will be on systematic methods of
analyzing the above data using statistical methods. The analysis of
the problems is not approached from the viewpoint of evaluating
the performance of indicator  parameters used in detecting some
type of change in groundwater quality but from the viewpoint of
understanding variability relevant to groundwater monitoring. To
improve  the  reader's  background on  analyzing variability,  the
paper begins by introducing some statistical concepts via an exam-
ple for separating variability associated with groundwater monitor-
ing. A simple procedure illustrated will test for differences among
concentrations of upgradient and downgradient wells.
Background of Groundwater Monitoring

  Presently, the Federal regulations promulgated under the re-
quirements of RCRA specify the use of a minimum of one upgra-
dient well and three downgradient wells to detect groundwater con-
tamination at hazardous  waste  sites. The upgradient  well  is
monitored  for one year  on a quarterly basis to  establish the
background concentration  of the four  indicators. The downgra-
dient concentration obtained from one of  the monitoring  wells
subsequent to the first year monitoring of the background well is
compared  to the background concentration  using Student's  t-test
or its modification.
  One  drawback of this approach is the wide ranging variability
associated with groundwater monitoring data, which creates prob-
lems in performing the t-test. It is generally felt that in order to ob-
tain good  quality data and perform a valid statistical test  it  is
necessary  to perform additional  steps of determining different
types of variation affecting groundwater quality.
  The results of  studies conducted by contractors under contract
with the USEPA,M some industry analysis2 and available ground-
water monitoring data3 indicate that background concentrations in
groundwater vary from season to season, from year to year and
from well to well. In addition, it is known that the results of sample
analysis vary from sample  to sample  and  analysis to analysis.
Although the sources of variance are extremely complex, major
groupings are:
•Temporal variation (including between-year and between-
 season variation) which could possibly be due to unsteady state
 plume migration and other effects
•Well-to-well variation which has a fixed component and a moving
 one
                                                     •Analytical variation
                                                     •Sampling variation
                                                       When groundwater is contaminated, the contaminants will be
                                                     transported downgradient of the area of contamination by ground-
                                                     water flow.  Dispersion and  retardation  will  have an affect of
                                                     diluting and attenuating the plume concentrations. If the ground-
                                                     water seepage velocity is small and the upgradient monitoring well
                                                     is sufficiently close to the  hazardous waste site, dispersion  pro-
                                                     pagating in symmetric directions will impact upgradient as well as
                                                     downgradient wells. The extent of the impact  will also be depen-
                                                     dent upon the magnitude of dispersivities prevailing in a particular
                                                     hydrogeologic setting.
                                                       The  plume concentration,  different  for different parameters,
                                                     may move over time or be  stationary; moreover,  it may be influ-
                                                     enced by  geologic material  and  well  construction differences.
                                                     Seasonally may be infiltration related, so dates are less important
                                                     than rainfall (e.g., inches of rainfall since the last frost of spring) or
                                                     it may  be temperature related. However,  complexity as always is
                                                     only resolved by simplifying assumptions. In this paper, the author
                                                     has divided variance into four sources, but any of the four could
                                                     similarly be subdivided.
                                                       In a  part of this paper, the author discusses in more detail how
                                                     these variables affect the observed results of groundwater quality
                                                     and  how groundwater contamination can be statistically detected
                                                     given the variability mentioned above. Before proceeding with the
                                                     investigation, one could raise several interesting questions. Among
                                                     those questions would be "How many data points does one need to
                                                     establish  the  reliable  background concentration  of  upgradient
                                                     wells?"
                                                       The  efficiency of monitoring study can be  increased by  using
                                                     statistically designed  well-sampling schemes  (called  designs by
                                                     statisticians).  Engineers are  concerned  with  well-monitoring
                                                     schemes and  the analysis of data. Well  sampling  combinations
                                                     should  be planned so that all relevant information can be efficiently
                                                     extracted from the data. The design of a well  sampling scheme is
                                                     more important than  analysis, because without a good design, it
                                                     would not be possible to obtain so much information with the same
                                                     amount of work. The author begins with some actual examples to
                                                     illustrate the basic simplicity of well-monitoring design.
                                                       The  following example, borrowed from the concept of two level
                                                     factorial designs, shows how 16 well-planned  monitoring samples
                                                     can be used to factor out seasonal and analytical variability and to
                                                     reach reliable conclusions on groundwater contamination. These 16
                                                     monitoring data  points represent the  two level three variable fac-
                                                     torial design. Where illustration  is needed, the field data will be
                                                     used wherever  possible.  The lack  of actual  monitoring  data
                                                     hampers illustration based on the field data in certain cases. This is
                                                     especially true in demonstrating the sampling variability among the
                                                     four variables.
                                                       The  examples given below use the field data to demonstrate the
                                                     well-to-well, seasonal and analytical variability and can be extended
                                                     to determine the effect of sampling variability. This may not be
                                                     necessary if the samples are randomly taken and if the objective of
                                                     the statistical test is only to find the statistically significant increase
                                                     between the upgradient and monitoring wells.  Because of the lack
                                                     of field data, the examples involving the three variables (well-to-
346
POST CLOSURE

-------
well, seasonal and analytical) use field data for one level of sam-
pling and hypothetical data for the other level of sampling.
A WELL SAMPLING EXAMPLE

  To  facilitate understanding of separation of the effect of each
variable on the concentration change, a well sampling scheme using
a factorial  design involving three variables (well-to-well, seasonal
and analytical) is used. Two levels of well sampling  are used to
establish the background concentration. It is assumed that there are
two upgradient wells identified as Well #1 and Well #2. These two
background wells are sampled at two different seasons, identified
as Season #1  and Season  #2. For example,  Season #\ may  cor-
respond to the 2nd  quarter.
  To study the effect of analytical variability at the two levels, a
sample is split into two for analysis. The split  samples will be
designated  as  Split  # and Split #2. The same approach can be ex-
tended to other cases where one upgradient and  more than two
background wells are used.
  Performing all possible combinations of well sampling to accom-
modate these  variability requires 23 = 8 analyses. To  increase the
precision of the results and to obtain an estimate of random error,
these eight sets of monitoring samples can be repeated. Hence, two
replicate analyses are  performed on each  split sample. Thus, in
reality,   obtaining  split  samples  for  analytical  variability  and
replicate analysis of each split sample amounts to performing  four
replicate analyses on each well  sample. However, each replicate
analysis is assumed  separated from each split sample. This gives 16
data points in all.
Well Sampling Conditions

  As described previously,  the three factors affecting variability are
selected for illustration. These include:  (1) well-to-well, (2) seasonal
and (3) analytical variability. For each factor, two levels chosen are
identified by Level tt\ and Level #2. The variability matrix is shown
in Table 1.
                                           Notice that no two rows have the same sampling conditions. This
                                           array can be thought of as merely a collection of eight different
                                           samples. For example, the actual sampling conditions for Test No.
                                           1 are upgradient of Well #1, Season #1 (or 1st quarter) and Split tt\.
                                           The experimental conditions for Test No. 6 are upgradient Well #2,
                                           Season #\ and Split #2.

                                           Groundwater Monitoring Data

                                             According to the  monitoring scheme, the sampling from the
                                           upgradient wells can be performed. To duplicate each of the eight
                                           samples, a split sample is replicated. This process involves duplicate
                                           analysis of a split sample. Since  each of the split samples from two
                                           different upgradient wells of two different seasons is repeated, 16
                                           data points in all are obtained. The purpose of the replicate analysis
                                           is two-fold: (1)  to increase the precision of the answers and (2) to
                                           provide an estimate of the intrinsic variation in the sampling and
                                           analysis method.

                                                                      Table 3
                                                   Analysis Results of Samples from Upgradient Wells
Test No.
1
2
3
4
5
6
7
8
Well-to-Well
Well 11
Well 12
Well «1
Well 12
Well »1
Well «2
Well 11
Well 12
Seasonal
Season 11
Season 11
Season 12
Season 12
Season 11
Season 11
Season 12
Season 12
Analytical
Split »1
Split 11
Split 11
Split 11
Split 12
Split »2
Split «2
Split 12
TOX Cone, (ug/1)
Ca, CjjlAv.Conc.
43, 42 42.5
46, 32 39
34, 29 31.5
33, 27 30
43, 25 34
35, 36 35.5
27, 27 27
37, 28 32.5
                            Table 1
                 Factors Affecting Data Variability
                    (two levels of three factors)
                                                                      Table 4
                                                   Analysis Results of Samples from Downgradient Well
 Factors
 well-to-well
 Seasonal
 Analytical
Level 1
Upgradient Well #\
Season #1
Split #1
Level 2
Upgradient Well #2
Season #2
Split #2
 Samples taken at each level of the three factors are replicated.
 Well Monitoring Scheme

   Eight sampling includes all combinations of two levels of three
 factors. The eight combinations of well sampling are given by the
 eight rows of Table 2.
                            Table 2
            Well Sampling Based on a 23 Factorial Design
Test No.

1
2
3
4
5
6
7
8
Well-to-Well
W
Well 1
Well 2
Well 1
Well 2
Well 1
Well 2
Well 2
Well 2
Seasona
(#)
Season
Season
Season .
Season .
Season
Season
Season
Season ;
Analytical
W
Split 1
Split 1
Split 1
Split 1
Split 2
Split 2
Split 2
I Split 2
Season
Season #1
Season #2
TOX Cone, (/ig/l)
1200
1000
                                             Sixteen observed values of groundwater sample analysis for TOX
                                           are given in Table 3. The data were retrieved from STORET and
                                           are the actual field monitoring data taken on May 19,  1982, and
                                           Aug. 12, 1982 for Season #1 and Season #2 samples, respectively.
                                           The last column of the table shows the average concentration of the
                                           replicate analysis for each of the eight distinct well samples.
                                             Analytical results of samples taken at two different seasons from
                                           a downgradient well were also  retrieved from STORET and are
                                           shown in Table 4. These data will be used for a statistical test later.
                                             So far,  the sampling variability has not been  considered. The
                                           reason for deferring the inclusion of the sampling variability relates
                                           to its difficulty in representing the affect of the four variables.
                                           Another reason for not discussing it earlier  is  the unavailability of
                                           the field data taken at  different levels of design for the sampling
                                           variability. The different levels of sampling are represented as Sam-
                                           ple #1 and Sample #2. All conditions of data variability can be writ-
                                           ten down as a combination of the two levels of the four variables.
                                             The 16  distinct  combinations of the  sampling conditions are
                                           given by the 16 rows in Table 5.  This  array in Table 5 can be
                                           thought  of as merely a collection of 16 different test conditions for
                                           obtaining groundwater monitoring data. For example,  the actual
                                           conditions for Test No. 1 are Well #1, Season #1 and Split #1 from
                                           Sample #1. Similarly, the test conditions for Test No. 7 are Well #1,
                                           Season #2 and Split #2 from Sample tt\. Since all the test  conditions
                                           are  repeated, the well monitoring combinations for  the  four
                                           variables generate 32 data points in all.
                                                                                                    POST CLOSURE
                                                                                                                            347

-------
                            Table S
          Well Monitoring Combinations for Four Variables
Test

1
2
3
4
5
'
7
e
9
10
11
12
1)

M
IS
16



We 11 11
Wo M 12
Wo 1 1 •!
Wo 1 1 12
wo 1 1 11
We 11 12
We 1 1 11
We 11 12
We 1 1 11
We 1 1 12
we 1 1 I 1
We 1 1
We 1 1 il

Well 12
well 11
We 1 1 >2



Seal on 11
Season 11
soj-,on >2
sojson 12
season 11
be json * 1
SKason 12
Season 12
Season 1 1
Season I 1
se-1'.on 12
So.i',on 12
Season * 1

season •)
Season 12
S-.i-on 12


*
Split 11
Spl I t 11
Spl 1 t 11
Split il
Split t:
split 12
Split 12
Spin 1.'
Split 11
Spl 1 t 11
Spl It 11
Split II
Spill 1.'

Spl,. 12
Split 12
Spl It 12



sample 1 1
Sample 1 1
Samplo 1 1
sample 1 1
Sample 11
•.,lm,,le 1 1
Samplo 11
Sample 11
S.impl. 12
sjmp>- 12
sampl« 12
sam;,l« 12
..., 14
Ji 36 li.i
27 21 21
37 20 32.1
40 41 40. *.
47 11 39
It 14 1
33 21 11
44 30 37. !,

30 >1 )0.1
29 21 21
)7 17 31

Analysis of the Analytical Variability

  It is  now  appropriate  to investigate the significance of the
analytical variability. To design two levels of data collection for the
analytical variability, a sample taken at different levels of the well-
to-well and seasonal variation is split into two for analysis. Also,
recall that a split sample is analyzed twice to increase the precision
and to provide the intrinsic variability in the groundwater monitor-
ing method. In actuality, this requirement resulted in  analyzing a
groundwater  sample taken  from an  upgradient  well  in  four
replicates.
  For example, Test No. 1 and Test No. 5 use the same sample ob-
tained from Well #1, Season #1 and Sample #1. But they are split
into two, and replicated in analysis. The similar argument can be
made for  Test No. 2 and  Test No.  6,  etc. The effect of the
analytical variability on the results of groundwater monitoring data
is nothing more than an estimate of random error one is  trying to
determine  as  the  intrinsic variability.  Hence,  the analytical
variability can be included as part of the intrinsic variability in the
groundwater  monitoring  method. It  follows that one can in-
vestigate the effect of the three variables instead of four, treating
the analytical variability as part of the  intrinsic variability.
  The three variables of concern are the well-to-well, seasonal and
sampling variabilities.  If  one  is  concerned with the effect  of
analytical variability, a sampling design can be constructed accord-
ing to the  levels  shown  in  Table 5.  However,  if the analytical
variability  is  treated as part of the intrinsic variability,  the well
monitoring combinations will be reduced to eight distinct  test con-
ditions.
  The use  of the  data in Table 5 for the eight test conditions will
yield  four  replicate analyses for a sample. That  many  replicates
may not be  necessary to  provide an estimate  of the  intrinsic
variability.  Two replicates are sufficient for this purpose.
Hence, the well monitoring scheme based on two levels of the three
variables with two replicates will yield  16 data points in all.

Average Variabilty Effects

  All possible combinations of two levels of three variables requir-
ing eight well samples are shown in Table 6. The  table also shows
the actual monitoring data for Tests No. 1 to 4 and the hypothetical
test data for Tests No. 5  to  8.

Effect of the Well-to- Well Variability
  When concentrations between two upgradient wells monitored at
the same time are compared, there could be a significant change in
                                                         concentration because of the well-to-well and sampling variability.
                                                         This variability, in the absence of contamination, may be explained
                                                         by a different mechanism than occurs in the presence of contamina-
                                                         tion.  How  would  one consider this  well-to-well variation and
                                                         evaluate its  influence on the change of concentration between the
                                                         upgradient and downgradient wells in the process of statistical com-
                                                         parison?
                                                           Before answering this question, one should investigate the range
                                                         of concentration change between two different wells in the absence
                                                         of contamination.
                                                           One should note that the sampling and seasonal conditions for
                                                         Test No. 1 and Test No. 2 are the same,  but the samples are ob-
                                                         tained from  different wells. Well #\  is used for Test No. 1 and Well
                                                         tn. is used for Test No. 2. Therefore,  the difference of concentra-
                                                         tion in this pair of tests, apart from intrinsic variability present, can
                                                         be attributed solely to the effect of the well-to-well variation alone.
                                                         Similarly,  for the pairs of Tests No. 3 and 4, 5 and 6, and 7 and 8,
                                                         each pair involves similar sampling conditions with respect to the
                                                         sample and season, but different sampling conditions with respect
                                                         to wells. Thus the  difference between each pair of the  results
                                                         reflects the effect of the well-to-well variability alone.
                                                           The differences of concentration for the example shown are
                                                         -3.5( = 39-42.5),  - I.5( = 30-31.5),   -1.5( = 39-40.5)  and
                                                         -4( = 31 = 35). The overall average effect of the well-to-well varia-
                                                         tion, Ew, is  the average of all four  differences:
                                                            Eu- -3.5-1.5-1.5-4.5
                             -2.75 ug/1
(I)
                                                         The formula  for calculating the average effect of the well-to-well
                                                         variability is:
                                                            EW- - (c2-cr:4-c3*c6-Cs-:a-c7l
                                                              =• - [39-42.5 + 30-31.5*39-40.5*31-35]
                                                                4
                                                                -2.75  ug/1
                                                           (2)
                                                         where C|..., Cg are the concentrations obtained for Test Nos. 1...,
                                                         8 respectively.
                                                           The average effect of the well-to-well variability is equivalent to
                                                         the main  effect of the well-to-well variability, a  common  ter-
                                                         minology utilized in the field of factorial designs. The result is the
                                                         observed effects of the well-to-well variation averaged over all the
                                                         other combinations of the other variables. It is the effect of chang-
                                                         ing the well from one to another which results  in the change in con-
                                                         centration.

                                                         Calculation of Confidence Intervals

                                                           If one obtains an answer of 2.75 /ig/1 for the average effect of the
                                                         well-to-well  variation, this means that when measurements are
                                                         made from one well to another, the concentration has changed by
                                                         2.75 /tg/1 on the average in the absence of contamination. It maybe
                                                         that 95% confidence interval is 2.75 ± 0.5,  but on the other hand,
                                                         the confidence interval may be 2.75 ± 50. There is a difference in
                                                         these two cases.
                                                           95% confidence intervals can be determined using the following
                                                         equation:
                                                             statistic ± t
                 /estimated variance of the statistic           (3)
where the value of t is the critical value read from a t-table, and the
statistics of interest  are the average effect of the well-to-well varia-
tion.
  The estimated variance of the statistic can be determined from
knowledge of the intrinsic variability of the monitoring procedure.
As indicated before, if the 95"% confidence interval were 2.75 ±
0.5, one would feel that an average variation of concentration be-
tween wells  is rather  convincingly demonstrated,  and one would
348
POST CLOSURE

-------
                           Table 6
        Well Monitoring Combinations for Three Variables
Test No.
1
2
3
4
5
6
7
1 8
We 11- to- We 11
Well *1
Well »2
Well »1
Well »2
Well HI
Well »2
Well »1
We 1 1 1 2
Seasonal
Season »1
Season 11
Season «2
Season 12
Season HI
Season HI
Season #2
Season ft 2
Sampl ing
Sample #1
Sample #1
Sample 111
Sample 11
Sample *2
Sample 12
Sample I 2
Sample 112
TOX (ug/1)
Ca Cb Av Con.
43 42 42.5
46 32 39
34 29 31.5
33 27 30
40 41 40.5
47 31 39
36 34 35
33 29 31
assert that the well-to-well variation would be fairly close to 2.75. If
the interval were 2.75 ± 50, this would not be the case at all.
  To obtain a quantiative measure of the uncertainty in the average
effect of the well-to-well variation, one calculates the within-groups
mean square by pooling  all estimates of a common variance a2.
Returning to the example in Table  6,  one can calculate eight
estimates of the variance s^, s^,—, s*, one for each test. In this ex-
ample, the degree of freedom k = 8 is determined from k = a = b = c
(n-1) where a is the number of test for the well-to-well variation, b
is the number of test for the seasonal variation, c  is the number of
test for the sampling variability and n is the number of replicates
per cell (n = 2 in the example).
  The  calculation of s^, the estimate of a2 from the two observa-
tions under conditions of Test No. 1 can be made as follows:
         (43-42.5)2 + (42-4.5)2  = 0.5
                                                        One is 95% confident, therefore, that the average of the well-to-
                                                      well variation is between -5.776 /tg/1 and 11.276 fig/I. In other
                                                      words, it is unlikely that the true well-to-well variation is lower than
                                                      - 5.776 /tg/1 and higher than 11.276 /tg/1.
                                                      Determination of Statistically Significant Increase

                                                        One uses the data in Tables  4  and6 to obtain  the difference in
                                                      concentration between the upgradient and monitoring wells.  In
                                                      practice, it is entirely possible that the sampling variation may be
                                                      included as part of the intrinsic variability. Since no data are
                                                      available to determine the sampling variability, one can calculate
                                                      the  difference in  concentration between the  upgradient and
                                                      monitoring wells based on Sample # 1 and  Sample #2.
                                                        For Season #1, the concentration difference is
                                                           1200 - 42.5 =  1157.5, 1200 - 39 =  1161
                                                           1200 - 40.5 =  1159.5, 1200 - 39 =  1161

                                                        For Season #2, the concentration difference is
                                                           1000 - 31.5 =  968.5, 1000  - 30 = 970
                                                           1000 - 35 =  965,  1000  - 31  = 969.
                                                        The average difference in concentration is
                                                        1157.5 + 1159.5 + 1161 + 1161 + 986.5 + 965 + 970 + 969
                                                                                                                        1063.4 w/l
                                                                     The difference thus determined exceeds the 95% confidence in-
                                                                  tervals for the well-to-well variation. In addition, the individual dif-
                                                                  ference in concentration between the upgradient and downgradient
                                                                  wells exceeds the effect of the well-to-well variation which lies bet-
                                                                  ween - 5.776 /ig/1 and 11.276 /tg/1 in the example. Hence, the data
                                                                  show a statistically significant increase.
Proceeding in the same way for Tests No. 2, 3111,8,  one gets

insert line that I can't set.

Hence the pooled variance Sp2 equals:
     c ,    0.5 + 98 + 12.5 + 18 + 0.5 + 128 + 18 + 162  = 54.69
     SP2  =  - - -


  Consider a statistic which is a linear combination of the observa-
tions, where
                                                           (4)
statistic =
                           + a2c2 + . . . + ancn
where one has n observations ci,C2,...cn, and a's are constants. The
variance of the statistic V(statistics) is given by:
    V(statistic) = (a  +
                                               (5)
  If the variance of each of the 16 analyses C in Table 6 is a2 and
the 16 analyses are uncorrelated, Vw, the variance of the average ef-
fect of the well-to-well variation is
                       cla
                       C8a
                                         C8b ~ C7a   C7b> '
                                                           (6)
  Since the true variance a2 is not known, it will have to be replaced
by Sp2. The t-value for a 95%  confidence region based on eight
degrees of freedom is 2.306. The estimated variance of the statistic
is Sp2/4 or Vw = 54.69/4 = 13.67. Thus the 95% confidence inter-
vals for the well-to-well variation are:
                   2.75 ± 2.306    /13~6T
                        = 2.75 ± 8.526
APPLICATION OF THE METHOD TO
OTHER SITUATIONS
  The method presented above makes use of the concept of fac-
torial designs and the analysis of variance. The method can be ex-
tended to the cases where data are obtained under conditions dif-
ferent than the example. These cases are as follows:
•There are three downgradient wells and one upgradient well. In
 this  case, the well-to-well variance and its confidence intervals
 will be determined from the three downgradient wells. The con-
 fidence intervals will be compared with the difference in concen-
 tration between the upgradient and monitoring wells.
•There are one upgradient and three downgradient wells as in the
 above case, but the quarterly measurement for one year will in-
 clude the downgradient as well as upgradient wells to establish the
 background well-to-well variance.  Once the well-to-well variance
 is estimated,  the  concentration difference  between the upgradi-
 ent and downgradient wells monitored after  the facility began
 operation can be  compared as above.
•There are two (or more) upgradient wells and three downgradi-
 ent wells. Data coveing more than two seasons can be used to in-
 crease the accuracy of the results.
                                                      REFERENCES

                                                      1. JRB  Corp., Evaluation of Statistical Procedures for Groundwater
                                                         Monitoring, prepared for the  USEPA,  Washington,  DC, Dec. 22,
                                                         1983.
                                                      2. Monitoring data sent to the USEPA Region by Phillips Petroleum
                                                         Co. as part of permit application.

                                                      3. Groundwater monitoring  data from Texas  and  South  Carolina,
                                                         stored in STORET.

                                                      4. Radian Corp., Statistical  Techniques for Evaluation  of Monitoring
                                                         Data in a Complex Groundwater System at LeHillier, MN,, prepared
                                                         for the USEPA, MERL, Cincinnati, OH, Apr., 1984.
                                                                                                     POST CLOSURE
                                                                                                                349

-------
     NATURAL  RESOURCE  RESTORATION/RECLAMATION
                            AT  HAZARDOUS WASTE  SITES

                                         JAMES R. NEWMAN, Ph.D.
                                            DOUGLAS S.  McLEOD
                                 Environmental Science and Engineering, Inc.
                                               Gainesville, Florida
                                            JACKSON B. SOSEBEE
                                 Environmental Science and Engineering, Inc.
                                             Englewood,  California
INTRODUCTION
  The primary emphasis in hazardous waste management has been
the protection of human health through site cleanup or site restora-
tion. Restoration to protect ecological health or natural resource
values has been given minimal attention. CERCLA addresses "the
potential for destruction of sensitive ecosystems" (Section 105) and
"injury to,  destruction of, or loss of natural resources" (Section
111) and (1) charges federal officials (i.e., Secretaries of Defense,
Interior,  Agriculture and  Commerce) through Executive Order
12316,  as trustees for  natural  resources  and (2) designates  the
USEPA as responsible for assessing such damages  and developing
action plans for recovery. In this paper, the authors discuss the re-
quirement for natural resource  damage and restoration/reclama-
tion  assessments of hazardous waste sites  and consider the appli-
cability of existing ecological and economic assessment techniques
to address damage and restoration/reclamation of hazardous waste
sites.
STATUS OF NATURAL RESOURCE DAMAGE
ADDESSMENT AND RESTORATION
  Estimates of hazardous wastes sites in the United States range
from 4,800 to 50,000.' By the end of 1983, 539 sites had been clas-
sified by the USEPA as Superfund Sites, meaning these sites pose
imminent danger to human health and the environment.2 A  na-
tionwide survey' to characterize hazardous waste sites has revealed
that most of the damage or contamination to ecosystems and other
natural resources is due to leachate, leaks or spills. The air, ground-
water, surface water and soil are the environmental media most
often affected.  Of 375 hazardous  waste sites surveyed (for which
there was sufficient  information),  30%  had  damage to  biotic
natural resources including flora, fauna and food chains.
  Although this survey identified a significant percentage of haz-
ardous waste sites as having or potentially having natural resource
damage, neither the 1982 nor the 1983 report of the Council of En-
vironmental Quality Reports mentions natural resource damage as
an issue.  In  the  1982 report,  the  natural resource  sections of
CERCLA are  not considered  major  provisions of  CERCLA.'
Recently,  the Office  of Technology Assessment published its 3-
year study on  the technology  options for managing hazardous
wastes,  means to address the problem of uncontrolled and  aban-
doned hazardous waste sites and the  adequacy of the federal regu-
latory program. No mention is made  in that 400-page document of
natural resource damage and the regulatory requirement to address
the issue.'
                                                      Sections 301(c)(l) and (2) of the CERCLA require the federal
                                                    government to promulgate regulations for the assessment of dam-
                                                    ages for injury, destruction or loss of natural resources resulting
                                                    from the release of oil or hazardous substances. These regulations
                                                    were to be promulgated no later than Dec.  12, 1982. Two types of
                                                    regulations are specified. Type A  regulations are to "specify stan-
                                                    dard procedures for simplified assessments  requiring minimal field
                                                    observation, including established measures of damages based on
                                                    units of discharge or release or units of affected area." Type B  reg-
                                                    ulations involve the development of regulations for conducting
                                                    assessments in individual cases. Specifically,  these procedures are
                                                    to determine  the type and the extent of short- and long-term in-
                                                    jury, destruction or loss of natural resources. Type B assessments
                                                    must include  replacement value,  use value and the ability of the
                                                    ecosystem or  resource to recover. Section  301(c)(3) provides that
                                                    these regulations be reviewed and  revised every two years. To date,
                                                    no such regulations have been promulgated.
                                                      Recently, the USEPA and the Department  of Interior (DOI)
                                                    have become  defendants in several law suits  regarding natural re-
                                                    source damage and the lack of natural resource damage assessment
                                                    regulations (e.g.. Civil Action No. 84-1668,  U.S. District Court for
                                                    the District of New Jersey). The Department of Interior is cur-
                                                    rently developing such regulations. On Jan.  10, 1983, the DOI pub-
                                                    lished a public notice in the Federal Register1 indicating its inten-
                                                    tion to develop such regulations  and to request information  and
                                                    suggestions for developing such regulations. On Aug. 1, 1984, the
                                                    DOI published  a  second notice  to inform the public about  the
                                                    nature of the  responses to the first notice.'  Currently, a work plan
                                                    tor natural resource damage assessment regulations is being devel-
                                                    oped. A number of the agencies in the DOI (e.g.,  Bureau of Land
                                                    Management) are also developing specific policies.
                                                      CERCLA  also  provides for law suits to  recover damages to
                                                    natural resources.  Several states have attempted to recover alleged
                                                    losses to natural resources under  CERCLA.  This is best exempli-
                                                    fied by suits involving Rocky Mountain Arsenal  in Denver, Col-
                                                    orado. The State of Colorado has sued the United States,  the
                                                    United States  Army, Shell Oil Company and Shell Chemical Com-
                                                    pany; the United States has, in turn, sued Shell Oil Company.
                                                      In the suit brought by the  United States  against Shell Oil Com-
                                                    pany, one of the claims for relief states that, "As a result of the re-
                                                    leases and threatened releases of Shell's chemicals at the Arsenal,
                                                    natural resources on, over and under the  Arsenal;  including  air,
                                                    land, birds, fish, wildlife,  biota,  lakes and other surface waters;
                                                    and groundwater,  belonging  to and held in trust by the Secretary
                                                    of the  Defense, have been and continue to be injured, destroyed
350
POST CLOSURE

-------
or lost." The Secretary of the Army is conducting an assessment of
the damages to such natural resources and currently estimates  the
amount of such damages to be approximately $1.8 billion, includ-
ing damages for harm to wildlife and the cost of decontamination
of natural resources contaminated with Shell's chemicals.
  Stein et a/.,7  in a survey  of procedures for determining costs of
natural resource damage from oil or hazardous wastes, provide a
preliminary view of the status of natural resource damage assess-
ment at the state level. As of 1983, a few states (such as California,
Florida and  Washington) had attempted  to conduct  detailed
natural resource damage assessments for several hazardous waste
sites or spills. Eleven other states have  conducted partial evalua-
tions of destroyed or  damaged natural resources. Most of these
assessments have been associated with oil spills or "fish  kill" in-
cidents.
  The authors believe that most damage assessments are conducted
on an  ad hoc basis as conditions warrant and agency  funds per-
mit. The focus of these assessments has been on highly  visible and
immediate damages such as direct mortality rather than on consid-
eration of inherent or long-term injury or loss to wildlife habitats
or ecosystems. The assessment approaches and  procedures used
are procedures based on existing oil and water  pollution regula-
tions.
  The lack of attention to natural resource damage considerations
appears to follow a pattern  found in man's attempt to control new-
ly perceived pollution problems. The initial efforts are  directed at
correcting immediate human health concerns (e.g., Love Canal
and Times Beach)  with secondary attention given to  natural re-
source  considerations. Initially, air  pollution  was seen primarily
as a human health problem; however, as the full extent of air pollu-
tion effects became understood, both human health and  environ-
mental effects were investigated.8
NATURAL RESOURCE DAMAGE
ASSESSMENT CONSIDERATIONS
  Natural resource damage,  as  set forth  in CERCLA  Section
 101(6), means injury to or loss of natural resources. For hazardous
waste sites, the causal  agent is a chemical substance which affects
the  natural  resource,  rendering it unusable because of possible
health effects or causes injury or death to the biotic components
of that resource.  The types of effects hazardous waste has on
natural resources is diagramatically shown in Figure 1.
  Basically, hazardous waste will cause both acute and chronic
effects to the biotic components of natural resources and will cause
contamination of the abiotic components. The contamination of
                        HAZARDOUS WASTES
 Acute Effects  to Biotic Resources:
 organisms, populations, communi-
Chronic  Effects to Biotic Resources:
organisms, populations, communities,
ecosystems
                        Contamination to
                       Abiotic Resources:
                        soil, water, air
                           Figure 1
     Conceptualized Diagram of the Effects of Hazardous Wastes on
                       Natural Resources
the abiotic component of the natural resource may cause addi-
tional acute and chronic effects to biotic  natural resources or
further contamination to other abiotic resources. In many cases,
acute effects will be short-term.
  At the Sapp Battery Superfund site in Florida, an acid waste dis-
charge initially caused death and injury to on-site and off-site wet-
lands. Fish kills also were reported until the operation was shut
down.9 Chronic effects  to natural resources are likely to be the
long-term characteristics of hazardous waste sites. At the Sapp Bat-
tery site, although the acid discharges have ceased, trace metal con-
tamination of the wetlands and upland areas has continued. Leach-
ing of these trace metals off-site has been observed.
  Comments  received by the DOI indicate that a definition of
natural resource damage is a first priority in developing the pro-
posed regulations.6  Based  on the general  policy guidelines of
CERCLA, discussions with  the Department of Interior and other
federal agencies and experience with hazardous waste site evalua-
tions and impact evaluations associated with "classical" pollution
problems, natural  resource damage assessment can be considered
to have three components: (1) affected natural resources and their
recognized values,  (2) damage assessment and (3) restoration.

AFFECTED NATURAL RESOURCES
  Section 101(16)  defines natural resources as "land, fish, wild-
life, biota, air, water, groundwater, drinking water supplies  and
other such resources belonging to, managed by, appertaining to
or otherwise controlled by the United States (or) any State." This is
a broad definition and can be interpreted to include almost every-
thing. Similar definitions were provided in other environmental
legislation [e.g., National Environmental Policy Act (NEPA)  and
the Clean Air Act]. These  general definitions are not  workable
definitions for assessment or management purposes.
  Controversies can arise between various interest groups as to
whether one or more "natural resources" are important and should
be  investigated. It is recommended that a  general definition of
natural  resources  be developed building on already-defined en-
vironmental concepts. It should be applied on a case-by-case basis
using already  identified natural resources  as a starting point for
specific damage assessment evaluations.
  Since the 1970s, considerable effort has been devoted to iden-
tifying important  environmental  resources  at the  federal, state
and local levels. For example, federal and state wild and scenic
rivers have been identified.  Federal and state endangered species
and their critical habitats are recognized. A national wetland class-
ification system has been developed as well as a system  of valua-
tion.10 Natural resource management plans exist for most govern-
mental land which still  has  natural environmental features. Fed-
eral, state and local land use plans (i.e., coastal management  and
comprehensive plans) also have  sections  devoted to  natural re-
source identification. For hazardous waste sites on or adjacent to
such lands, identified environmental resources can serve as a basis
for developing a list of natural resources for which damage assess-
ment procedures can be applied.
NATURAL RESOURCE DAMAGE

  In some ways, the definition of natural resource damage has the
same characteristics as NEPA's term "impact." Over the years, the
term impact has come to be understood as a significant change
from baseline environmental  and economic  conditions.  The en-
vironmental impact statement (EIS) literature includes discussions
of the meaning of impact or significant impact. Generally, "signif-
icant impacts" refer to a man-induced environmental change of
large magnitude on a resource and/or a change which affects an
important resource.
  The authors suggest that, for  regulatory  purposes, natural re-
source damage be defined as any significant injury or loss to  a
natural resource or any  significant loss of a natural resource value
as established by baseline conditions with the term significant im-
plying the same characteristics as in NEPA (i.e., being of large
magnitude and/or involving an important natural resource). Since
                                                                                                     POST CLOSURE
                                                                                                                            351

-------
human health considerations are paramount, significance should
also include  any natural resource  condition leading to  possible
human health effects.
  The National Oceanic and Atmospheric Administration recently
published  an  economic  damage  assessment   report"  on  the
AMOCO CADIZ oil spill. Although the spill was not a hazardous
waste  spill, the  authors recognized the  natural  resource damage
assessment requirements of CERCLA and attempted to relate their
experiences for conducting hazardous waste damage assessments.
Based on their experience, they identified several general categories
of problems  associated with natural resource damage assessment,
including:
•Uncertainties associated  with effects of spills on natural resources,
  in particular long-term effects of unquantified contamination or
  unmanifested damage
•Lack of data on preexisting conditions
•The lack of credible methodologies for making damage estimates
  for noncommercial resources (e.g., seabirds)
   Hazardous waste contamination is similar to the "classical"
kinds  of pollution (e.g.,  pesticide contamination, oil spills, acid
mine drainage and point and nonpoint source air and water pollu-
tion).  For these and other "classical"  kinds of pollution, hundreds
of studies, symposia and conferences as well as numerous clean-
up, restoration and pollution control programs  have been devel-
oped, tested and implemented.
   Some of  the major effects,  characteristics, assessment  guide-
lines and  regulatory standards  for several  "classical" pollution
problems which are similar to the kinds of problems found at haz-
ardous waste sites are summarized in Table  1. An extensive dam-
age assessment data base  has been developed to assess and control
these problems. A large number of natural resource agencies have
regulatory jurisdiction over these problems and could be a source
for applicable assessment  techniques.
                                                                   NATURAL RESOURCE RESTORATION/RECLAMATION

                                                                     As stated earlier, natural resource damage assessment also in-
                                                                   cludes a requirement for recovery of the damaged natural resource
                                                                   through restoration or reclamation. Recovery includes not only the
                                                                   recovery of a damaged or lost natural resource but also lost natural
                                                                   resource values.
                                                                     The goal of natural resource recovery is to return the natural re-
                                                                   source  to  the previous or  preexisting condition or value.  If this
                                                                   goal is not possible, the goal becomes reclamation of the site to an
                                                                   equivalent natural resource condition.
                                                                     Sustainable recovery programs should be included in the objec-
                                                                   tives  of restoration and reclamation.  There  should be a definite
                                                                   endpoint to restoration and reclamation so that continued  restor-
                                                                   ation and reclamation practices are not necessary.
                                                                     Depending on the kind of injury or loss and the costs for restor-
                                                                   ation and reclamation, the recovery process  can be by  artificial
                                                                   means (e.g.,  top soil removal and  replacement and revegetation)
                                                                   or natural means using  natural  ecological processes (e.g.,  biode-
                                                                   gradation) to "clean up" the contamination and to allow natural
                                                                   succession to revegetate the damaged or lost ecological community.
                                                                   Natural recovery  rates should also be  taken into account. Suc-
                                                                   cession proceeds rapidly in some natural resource systems (i.e., sub-
                                                                   tropical regions  like  Florida),  whereas  in  other systems (i.e.,
                                                                   deserts and alpine regions) recovery can be slow (e.g., decades to
                                                                   hundreds of years). Numerous restoration/reclamation procedures
                                                                   exist  for "classical" pollution  problems (Table 1) which are appli-
                                                                   cable at hazardous waste sites. The choice of restoration/reclama-
                                                                   tion procedures is a function of four conditions: (1) present con-
                                                                   dition of the site (i.e., type of natural resources and their values);
                                                                   (2) desired or required recovery rates; (3) desired land use alterna-
                                                                   tives; and (4) costs for  reclamation and  restoration. These con-
                                                                   ditions have to be evaluated on a site-by-site basis.
                                                                 Table 1
      Summary of Selected "Classical" Pollution Problems with Potential Applicability lo Natural Resources Damage and Restoration/Reclamation
                                                   Assessments of Hazardous Waste Sites
                           Characteristic!   Air Polltxion
                                                       Strip Hining
                                                                      Oil Spill.
                                                                                 Sctax* PAroff
                                                                                                 Appl tear ion
                           Onrral Cluut
                                  Gaseous, particulate
                                  inclining trace
                                  metall and acidi-
                                  fying millions
                Mechanical dittia
                bance and trace
                aetal and acid
                release
                                                                    Heavy Co lighc
                                                                    oil*
                           Major Reported  Injury and death to   Unproductive soils. Injury and
                       Natural
                       Resource
                       Effects
                                      flora and fauna;
                                      lake acidification
Host Connonly
Affected
Nstia-al
Resources
Mitigation
Alternat iwi

Vegetation, toils,
lake*


Enitsion control,
resistant species

Soil, vegetal
streams, rivei


Cantairnent •
treatment of
4«stes
                       Rest or at ion
                       Rpclamst ion
                       Program
                           Rrpretencac ive
                           Regulariona
Lak* liaung,        Howrous land
planting resistant    reclamation pro-
•peci*n           cedures, plans
                required for permit!
                injury and death   nation of fauna,
                to aquatic        and fauna,
                organissja, ground   surface niter
                and surface **ter   contamination
                contaninat ion
                                                                Aquatic
                                                                organianB,
                                                                sediawnts
Gone a invent,
collection, and
disposal of
contaminated
soils

Natural
recovery
            Urban and agricui- Organic and inorganic
            tural runoff—    coapound*
            organic and in-
            organic

            Surface and ground Toxicity to terres-
            Mter concauna-   nrul and aquatic
            lion, tOHicity to  organtaaa, food
            aquatic organissai  chain cantaunation
                                                                                              Soil, **(er ,
                                  Clean Air Act       Surface mi rung     Clean tatte
                                                  Control «nj RKla-  Act
                                                  nation Act; Clean
                                                  Water Act
                                                                                 Seraaaa., riv
                                                                                 lakes
                                                                                 Source controls—  Application of
                                                                                 structural and nan- controls, reduction/
                                                                                 structural       banning
                                                                                 Natural recovery  Natural recovery
                                                                                 dredging
                                                                                 Claan Uitar Act   FIFRA; TSCA
•NFS
 FWS
 FS
 EPA
              National Park Service.
              Fish and Wildlife Service.
              Forest Service.
Federal IK,
Retource
Agency
Information
Sourcea
ItCC.
iervicc.
al Protection Agency.
FVS. re, EPA




DM
FIFRA -
SCS
NRC
m, FVS, SCS KMA, EFA, CC EPA




Bureau of Mines.
Federal Insecticide, Fungicide, Rodencidc Act.
Soil Conservation Service.
Nuclear Regulatory Commission.





NOAA
CO
USDA
TSCA
IB», EPA, FS




- National Oceanic and Atmospheric Administration.
- Coast Guard
• U.S. Department of Agriculture.
- Toxic Substance Control Act .
                                                              Source. ESE, 1984.
 352       POST CLdsURE

-------
  Natural resource restoration/reclamation considerations for two
hazardous  waste sites are shown in Table 2.  At the Sapp Battery
site in Florida where cleanup is under way, the conditions are anal-
ogous to problems encountered with acid mine drainage from sur-
face mines. Groundwater contamination is  continuing.  Possible
effects on downstream surface waters and aquatic natural resources
are uncertain.  Restoration has emphasized groundwater restora-
tion.
  Although the courts have recognized that liability for natural re-
source damage includes losses to natural resources (i.e., wetlands
and their values),  wetland recovery is not a priority in the site
restoration  plans.  Wetland  reclamation  techniques successfully
used in Florida for phosphate mine reclamation are also applicable
for the Sapp Battery site.
  The Cordova Chemical site in Michigan has characteristics anal-
ogous to those found with misuse of agricultural pesticides. In this
case, the hazardous waste effects are not as complex; fewer natural
resources are affected,  and the full extent of the contamination
appears to be known. Two restoration/reclamation options were
recommended for the Cordova Chemical site: (1)  aquifer restora-
tion and (2) no action—allowing  natural recovery to occur (an
estimated 409 years).
  Techniques for aquifer restoration are well established. Relying
on natural recovery rates will result in the natural resource  (e.g.,
the stream) and its values being lost for more  than  a generation.  In
accordance with  CERCLA  policies, lost resources  and values
should be recovered sooner or replaced with equivalent resources
and values. In both these cases, several different natural resources
                             Table 2
 Examples of Natural Resource Restoration/Reclamation Considerations
                 at Selected Hazardous Waste Sites
   Characteristics
                      SAPP Battery Site
                        (Florida)
                                   CORDOVA Crmical Site
                                      (Michigan)
   Danaged Natural
   Resources
                SuLfuric acids and trace metals
                Known
                                          Orpanics
                                          Known
   Garage Natural
   Resource Values
    1. Soils
    2. Ground and surface water
    3. On- and offsite wetlands
    4. Offsite fisheries

    Potential
    1. Upland and on- and offsite
        wetland wildlife contamination
    2. Aquatic food web contamination

    1. All wetland functions
    2. Subsistence and recreational
        fishing
    3. Aesthetic values
                                          1. Soil
                                          2. Ground and surface water
                                          3. Sediment
                                          4. Offsite fisheries

                                          Potential
                                          1. Aquatic food web contamina-
                                             tion
1. Recreational fishing
2. Aesthetic values
  Similar 'Classical" Acid drainage fron mining
  Pollution Problwn

  Restoration/      Author ized/Recomiended
  Reclamation       I. Ground vater pulling
  Options         2. Collection of soils > lOOOppn
                   contamination
                              Agricultural pesticides contami-
                               nation

                              Author ized/Recoimended
                              1. Aquifer restoration involving
                                 ground water pumping and
                                 treatment
                              2. No action—allowing 40 yrs
                                 for natural recovery
                                (natural purge options)

Other Possible                  other Possible
I. Artificial restoration of onsite    None
   wetlands, allow natural recovery
   of offsite wetlands.

1. Degree of food web contamination    1. Ground water plume is moving
              unknown, monitoring needed, results
              could require greater soil
              decontamination
           2. Over S6.5 million for damages to
              natural resources awarded to
              Stare of Florida
                               to surface water stream; no
                               action option should include
                               intensive monitoring of
                               aquatic natural resources.
ppm = parts per million.
Source: ESE, 1984.
were damaged. Undocumented natural resource damage also is sus-
pected. Restoration and reclamation techniques used to treat sim-
ilar "classical" pollution problems are applicable to further sites.
  In one case (Sapp Battery site), natural resources restoration is
not presently being evaluated. At the  Cordova site, if natural re-
covery is allowed, a long-term damage assessment  program (i.e.,
monitoring) may be required.


ECONOMIC CONSIDERATIONS

  As previously stated, no specific guidelines exist  on the evalua-
tion of economic benefits and costs of natural resource damage
and restoration and reclamation assessments at hazardous waste
sites. Simplified procedures are requested in CERCLA. Procedures
currently used in a few states have attempted to develop replace-
ment values and measure indirect losses and use values. None of
the existing procedures meets the needs of CERCLA.7
Net Present Worth
  A net-present-worth approach is  recommended as the primary
economic technique to evaluate the monetary costs and benefits of
damage and recovery at hazardous waste sites. Nonmonetary costs
and benefits are best expressed in appropriate descriptive terms; a
six-step process is suggested for an economic evaluation:
•Determine which costs and benefits are applicable to the project
•Obtain project-related data on applicable costs and benefits
•Array those costs and benefits  that are nonquantifiable and/or
 cannot be evaluated monetarily
•Determine dollar value ranges for applicable costs and benefits
 and value ranges for key economic parameters
•Perform net-present-worth and sensitivity analyses
•Report economic findings
  Possible benefit and cost types affected  by a hazardous  waste
site are presented in Column 1 of Table 3. In an economic  sense,
these types of costs and benefits may be grouped into four cate-
gories:
•Those exchanged in  well-functioning markets and quantifiable
 (e.g., restoration/reclamation costs and damage to agricultural
 crops)
•Those quantifiable and  measurable goods/inputs exchanged in
 less than perfect markets (e.g., public water supply)
•Those quantifiable and  measurable  goods/inputs  that  are not
 normally exchanged in any type of market (e.g., recreation benefit
 flows from natural areas); because they are quantifiable in some
 sense, proxies can be used for market values
•Those nonquantifiable and  unmeasurable benefits and  costs for
 which dollar values are nearly impossible to determine because of
 difficulty in quantitative description (e.g., the benefit flow from a
 scenic view)
Economic Categories

  Breaking down benefit and cost types into economic categories
provides a sound economic basis to evaluate hazardous waste sites.
In performing natural resource economic analyses at hazardous
waste sites, it is important to determine dollar value ranges of key
 input variables and then  to perform sensitivity analyses on those
variables. Although reasonable dollar values should be  obtained
for each economic category (e.g., restoration/reclamation costs),
dollar values are subject to  even greater variability in economic
categories 3 and 4, which  contain numerous natural resource con-
siderations. Sensitivity analysis allows the economic analyst to vary
economic input values and compare the resultant  economic im-
pacts.  Although oriented to water resource  projects,  the U.S.
Water Resources Council procedures may be useful in determin-
ing dollar value ranges.12
  To properly assess the economic value of any natural resource
unit or area (e.g., freshwater marsh), the benefits it provides over
an impacted area and  over time  must be determined. Whereas  in
the traditional economic  market sense, the value  of natural re-
sources such as wetlands is primarily a function of its potential for
                                                                                                          POST CLOSURE
                                                                                                                                  353

-------
                            Table 3
     Benefit/Cost Types, Economic Categories, and Example Analysis
                         Where Evaluated
                               Analysis Where Type Evaluated  (X)
                                                     Nol
                                          Exchanged Exchanged
                                Economic      in       in

Li fe Cyc le Cos 1 s
Feasibil ity Studies 1
Land Acquisition/Relocation I
Construction/Reclamation I
Operation and Maintenance* I
Public/Private Buildings 1
Unemployed Resources
Income 2
Employment 1
Human Health and Safety 3, 4
Water Supply
Surface Water 2 , 3
Ground Water 2,3
Water Quality 2, 3
Agriculture I
Forestry 1
Livestock 1
Biotic Communities
Wetlands 3
Uplands 3
Wildlife 3, 4
Air 3
Recreat ion 3,4
Aesthetics 3, 4



X x
X
X
X X
X X
X X

X X
X X
Xy
f.
X

X X
X X
X
X X
X X
X

X
X
X
X
X
X
X
 Source: ESE. 1984
development, its best economic use from a CERCLA policy view-
point may be in its past use where greater benefits were provided
to man than are provided in the currently damaged  state. There
are many service flows from such areas that are not traded in mar-
kets. For instance, wetlands may provide wildlife habitat for threat-
ened, endangered  and other species and substantial food produc-
tion sources for fish. Wetlands also interact closely with uplands.
As a result, many of these services are joint in nature.
  Using the four economic categories, cost-benefit analyses may be
conducted to meet the specific characteristics of the hazardous
waste site or meet  the desires of different groups such as decision-
makers or a local interest group. In every circumstance, a life-
cycle cost analysis should be conducted. Life-cycle cost analysis in-
volves expressing all significant project costs over a study period
for various alternatives in  equivalent dollars.  Project cost guide-
lines may be obtained from numerous sources, including the Amer-
ican Society of Civil Engineers, "Manual Number 45—Consulting
Engineer"." Life-cycle costing techniques also may be found in
numerous texts.14'15
  Cost-benefit analyses also may be conducted for each economic
category inclusive  of all previous categories. For instance, an ex-
changed-in-markets analysis (Table 3, Column 3) would  consist of
all quantifiable goods/inputs normally exchanged in the U.S. econ-
                                                          omy (economic categories 1 and 2, Table 3). Frequently, an analy-
                                                          sis for each economic category is not appropriate because the cost
                                                          and benefit types may overlap the economic categories or certain
                                                          categories may not exist. The economic analyst should view  these
                                                          economic categories and levels of analysis as evaluation guidelines
                                                          and evaluate individual sites according to specific characteristics.
                                                          However, the grouping of types of benefits and costs has been use-
                                                          ful to decision-makers because they place their own relative impor-
                                                          tance on the economic categories.

                                                          Economic Parameters

                                                            For an economic analysis, economic parameters  (i.e., discount
                                                          rate, study year,  study period) must be established. A hazardous
                                                          waste site will incur costs (i.e., future injury and loss and restora-
                                                          tion/reclamation  plans) for many years. In order to make sound
                                                          natural resource  damage assessment decisions, monetary values
                                                          must  be identified by  amount and time.  People generally prefer
                                                          present benefits  to future benefits  (e.g.,  obtaining money  now
                                                          rather than sometime  in the distant future) for various reasons.
                                                          Amounts in different  time periods  may be put into equivalent,
                                                          present-worth units by multiplying future amounts by a factor be-
                                                          coming progressively smaller  for the more distant time periods.
                                                          The discount rate is the time rate of  decrease  in this factor ex-
                                                          pressed in percent per time period (e.g., 7.0%/yr). The higher the
                                                          discount rate, the smaller  is  the discount factor in future time
                                                          periods. A low discount rate tends to favor alternatives with rela-
                                                          tively high  benefits  and/or low costs in  the  future relative to
                                                         alternatives with lower benefits and/or higher costs in the future.
                                                           Although not necessary, it is recommended that  the  economic
                                                         base year be the  year  the economic analysis begins. Subsequent
                                                         present-worth dollar values would be reported in the base year.  The
                                                         study period is the length  of  time chosen for  consideration  and
                                                         study of incremental costs and benefits in the economic  analysis.
                                                         The study period may vary substantially by site and type of project.
                                                         CONCLUSIONS

                                                           CERCLA specifically  requires the  protection of natural re-
                                                         sources and associated values, as well as assessments of natural re-
                                                         source damage and the restoration/reclamation of injured or  lost
                                                         natural  resources. Little emphasis has been given to natural re-
                                                         source damage in most hazardous waste evaluations. The federal
                                                         government  is behind  in promulgating natural  resource damage
                                                         assessment regulations. Only a few states have attempted to fully
                                                         consider the natural damage aspects  of hazardous waste sites, al-
                                                         though  injury and loss to  natural resources are documented to
                                                         occur at a significant number of hazardous waste sites.
                                                           A number of problems exist in developing the required regula-
                                                         tions including working definitions of  natural resources, damage
                                                         and recovery. Simplified procedures  for natural resource damage
                                                         assessment (including recovery) are also needed. A considerable
                                                         body of applicable information exists in "classical" pollution lit-
                                                         erature. This information  includes assessment  and restoration/
                                                         reclamation  techniques and procedures. This information should
                                                         be reviewed for its applicability to hazardous waste site assessments
                                                         and recovery plans.
                                                           Natural resource restoration/reclamation should have as its
                                                         objective the return of natural resources and associated  values to
                                                         their previous or equivalent  condition. Both  artificial and/or
                                                         natural recovery procedures can be employed.  A comprehensive
                                                         net-present-worth  economic analysis including  natural resource
                                                         effects as project-related life cycle costs (restoration/reclamation)
                                                         should be performed.

                                                         REFERENCES

                                                          1. Office of Technology Assessment,  Technologies and Management
                                                            Strategies  for Hazardous  Waste Control,  Office of  Technology
                                                            Assessment, Washington, D.C., OTA-M-1%, Mar., 1983.
                                                          2. The  Council  on  Environmental Quality,  Environmental  Quality
                                                            1983, 14th Annual Report, Superintendent of Documents, U.S. Gov-
                                                            ernment Printing Office, Washington, D.C., 1983.
354
POST CLOSURE

-------
3. USEPA, Assessment of Hazardous Waste Mismanagement Damage
  Case Histories, OLffice of Solid Waste and Emergency Response,
  USEPA, Washington, D.C., EPA/530-SW-84-002, April, 1984.
4. The Council on Environmental Quality, Environmental Quality 1982,
  13th Annual Report, Superintendent  of Documents, U.S. Govern-
  ment Printing Office, Washington, D.C., 1982.
5. Federal Register, 48, No. 6, Jan., 1983,  1084-1088.
6. Federal Register, 48, No. 6, Aug., 1983, 34768-34772.
7. Stein,  R.J., Gundlach, E.R., Covell,  S. and Hodge,  J., Summary
  and Analysis of State and Selected Foreign  Procedures for Determin-
  ing the Costs of Natural Resource Damages from Releases of Oil or
  Hazardous Materials, Research Planning Institute, Inc., Columbia,
  SC.
8. Stern, A.C., "History of Air Pollution Legislation in  the United
  States," JAPCA 32, 1982, 44-61.
9.  Florida Department of Environmental Regulation, Sapp Battery Site
  Miscellaneous Memoranda, Florida Department of Regulation, TAlla-
   hassee, FL, 1983.
10.  Tiner, Jr., R.W., Wetlands of the United States: Current Status and
    Recent Trends, U.S. Fish and Wildlife  Service, National Wetlands
    Inventory Habitat Resources, Superintendent of Documents,  Wash-
    ington, D.C., Mar., 1984.
11.  U.S. Department of Commerce, National Oceanic and Atmospheric
    Administration,  Rand National Ocean Service, Assessing the Social
    Costs of Oil Spills: The Amoco Cadiz Case Study, Washington, D.C.,
    July, 1983.
12.  Federal Register, Part IX, Water Resources Council, "Procedures
    for Evaluation of National Economic Development (NED) Benefits
    and Costs in Water Resources Planning  (Level  C), Final Rule," 44,
    No. 242, Dec., 1979.
13.  American Society of Civil Engineers, Manual Number 45, Consulting
    Engineer,  1981.
14.  Dell'Isola, A.J. and Kirk, S.A., Life Cycle Costing for Design Profes-
    sionals, McGraw-Hill Book Co., New York, NY, 1981.
15.  Grant, E.L. and  Ireson, W.G., Principles of Engineering Economy,
    4th Ed., Ronald Press Company, New York, NY, 1964.
                                                                                                           POST CLOSURE
                                                              355

-------
  ASSESSMENT OF  GROUNDWATER  CONTAMINATION AND
 REMEDIAL ACTION FOR  A  HAZARDOUS  WASTE FACILITY
                                  IN A  COAL MINE REGION

                                               MARK J. DOWIAK
                                              ANDREZEJ NAZAR
                                                 NUS Corporation
                                            Pittsburgh, Pennsylvania
 INTRODUCTION

   Southwestern Pennsylvania is a region of the country with abun-
 dant bituminous coal resources. Historically, these coal deposits,
 in conjunction with the extensive river system in the Upper Ohio
 Valley, provided a base for development of a major steel industry.
   Coal deposits in the region are flat-lying and have been mined by
 both underground and open-pit strip methods. The open-pit mines
 are of interest at the subject site. Coal seams up to 130 ft below the
 surface are mined today in Pennsylvania by the  open-pit method.
 Dragline excavators with 20 yd1 capacity are capable of cost-effec-
 tively removing rock  overburden for coal seams only 5 to 10 ft
 thick.  Present surface mining regulations require that mine oper-
 ators construct erosion and sediment control and acid drainage
 treatment facilities and reclaim the open strip cuts to their approx-
 imate original surface contour.
   Past stripping operations used smaller excavating equipment
 than modern day draglines, and  the depth of the open cuts was lim-
 ited to 50 to 60 ft. Coal seams were usually stripped along the crop-
 line where overburden was thinner. Parts of the coal seam with
 greater overburden  were  deep-mined.  Open-pit mines  and deep
 mines were usually both developed at the same coal seam. Many of
 the older mines were not subject to the regulatory requirements of
 today. Open-pit mine cuts up to 50 ft deep were  left open and un-
 reclaimed. Approximately 440,000 acres in Pennsylvania are class-
 ified as unreclaimed mined lands.
  Unreclaimed open-pit coal mines provided a convenient location
 for disposal of a variety of municipal and industrial wastes. Since
 the mines were in rural, unpopulated areas,  they usually had old
 coal haul roads running to them.
  Unreclaimed land was purchased at a low cost  by landfill devel-
 opers and was commonly used  for waste disposal. Backfilling of
 open cuts with wastes was actually seen as an environmental benefit
 by regulatory agencies because of the restoration of the land  sur-
 face.

 SITE HISTORY AND DESCRIPTION

  The site under study is in an area of abandoned open-pit  and
 underground coal mines in Beaver County, Pennsylvania (Fig. 1).
 The privately owned disposal facility began operations in 1959 as a
 waste service for the regional steel production and manufacturing
 industries. Approximately 2 to 8  million gal/month of mostly spent
 pickle liquors were processed, the majority of which were sulfuric
 acid. Operations ceased in 1982 under order of  the Pennsylvania
 Department of Environmental Resources (PADER).
  Waste  types  that  had been  accepted  at the  facility  included
 corrosives, sludges from electroplating operations, spent stripping
      OHIO
       PENNSYLVANIA

+ SITE LOCATION

   • PITTSBURGH
         W. VA.
                         Figure 1
                    General Site Location
and  cleaning bath  solutions  from electroplating  operations,
quenching sludges from metal heat  treating, spent pickle liquor
from steel finishing and sludges resulting from the lime treatment
of spent pickle liquors.
  The facility (Fig. 2) consisted of a waste receiving, storage and
neutralization area and a 45 acre disposal impoundment. The im-
poundment was constructed in an  abandoned  stripcut of the
Mahoning Coal. Containment of the wastes was  achieved by the
remnant stripcut highwall and the downslope retaining embank-
ments constructed of local mine spoils. No liners or leachate collec-
tion systems were installed.
  Raw wastes were received by tank truck and discharged to a lined
waste holding tank. The acid composition waste stream was neu-
tralized with a hydrated lime-water slurry to approximate pH 9.5
to 10.0 and discharged to the adjacent sludge disposal impound-
ment. This is the basic  waste treatment process used since 1959.
The facility is considered a "Hazardous Waste"  landfill because
lime-treated wastes are generically listed  as "hazardous" under 25
PA Code Section 75.261(h), February, 1981.
  For the last 2 years of the facility operations, a portion of the raw
waste stream was  processed  through a waste  stabilization pilot
plant. The process included addition of various ratios of bentonite
clay and Portland Cement to the neutralized wastes.
  After treatment by the solidification process, these wastes were
rendered "non-hazardous" under a delisting granted by PADER.
The solidification operation also had the potential of providing a
low permeability cover for the in-place lime sludge. This cover was
356
         POST CLOSURE

-------
evaluated as part of the closure alternatives study and is discussed
in a later section of the paper.
   The lime-neutralized  wastes were discharged into the disposal
impoundment and flowed by gravity to the lowest elevations of the
impoundment. The  solution was a semi-solid slurry of 5  to  10%
solids which freely settled and dewatered to a stiff sludge over time.
From 50 to 75 ft of sludge are present in the impoundment.
                                            WASTE  TREATMENT
                                            FACILITIES
                                     CROPLINE  OF UPPER
                                     FREEPORT COAL
                            Figure 2
                        Facility Plan View

WASTE CHARACTERISTICS
  Sludge physical properties and chemical leaching characteristics
are summarized in Tables 1 and 2, respectively. The sludge is gen-
erally a soft, wet, highly plastic silt soil with loose,  compressible
and sensitive structure. According to the Unified Soil Classifica-
tion System, the sludge is an MH soil, i.e., a silt of high plasticity.
The high plasticity and loose sensitive structure are inferred to re-
sult from the chemical composition of the sludge and its mode of
deposition. The sludge has low to very low permeability and poor
drainage characteristics. Because most of the sludge particles are of
silt  size,  capillarity tends to  hold water in the voids between par-
ticles, thereby  preventing gravity drainage. Capillary effects are
also responsible for drying,  desiccation and shrinkage of sludge at
the pond surfaces.  Wet sludge has  very low undrained shear
strength, while desiccated sludge has moderate undrained shear
strength. Under drained loading conditions, wet or dry sludge has
high frictional shear strength.
  Chemical characteristics shown in Table 2 indicate that contam-
inants from older sludges and site seepage are similar and are pri-
marily inorganic anions and ammonia nitrogen. Metals are below
the USEPA National  Interim Primary Drinking Water Standards
(NIPDWS). Organic priority pollutants were analyzed for and were
not detected at significant levels and are not reported here.
  The site seepage has greater concentrations of chemicals than the
lime-neutralized  sludge leachate.  The reasons  for the difference
are not clearly understood; however, several hypotheses have been
suggested:
•Variations in physical and chemical environment of sludges in situ
 and sludges in the EP Toxicity Test. Field leaching is at a much
 less dynamic level than the EP shake test and dilution effects are
 greater.
•The effects of the coal mine waste materials on the chemical char-
 acter of the sludge leachate. Mine spoils are acidic and leach sul-
 furic acid, anions, cations  and metals.  The reduced  pH  of the
 seepage is indicative of the mine waste effects.
•The lime-neutralized  sludge samples taken in the field for the EP
 Toxicity testing have dewatered from approximately 8 to 50%
 solids, by weight. The free  water which was lost prior to  taking
 the sample is probably more indicative of seepage quality than the
 EP Toxicity leachate.

INVESTIGATION METHODS
  Site investigations were conducted in two phases. These inves-
tigations included field reconnaissance, test borings, monitoring
well installations and groundwater and surface water sampling.
  Twenty-seven test borings  were made, and 46 observation wells
were constructed in the borings with each boring containing from
one to three  observation wells.  Multiple wells were installed  at
different elevations corresponding to the aquifer  interval. Open
boreholes were  left in the shallow  rock aquifers.  Permeability

                            Table 1
                 Waste Sludge Physical Properties
Waste Gravity of Solids, Gs
Water Content, w, %
Dry Density, lb/ftj
Atterburg Limits (w, %)
  Liquid, LL
  Plastic, PL
  Plasticity Index, PI
Grain Size Distribution
  Sand (larger than 0.074 mm),
  Silt (0.074 to 0.002 mm), %
  Clay (smaller than 0.002 mm),
Permeability, k, cm/sec
Consolidation Characteristics
  Initial void ratio, e
  Initial dry unit weight, lb/ft5
  Initial total unit weight, lb/ft3
  Initial saturation ratio, s, %
  Compression index, Cc
    3.41
  191
   35.7


  101
   81
   20

   11
   81
   20
3.2x10-'

    7.75
   24.3
   76.8
   94
    2.67
                            Table 2
       Sludge Chemical Leaching Characteristics and Average Site
                         Seepage Quality
Chemical Parameter
Barium
Cadmium
Mercury
Silver
Chromium
Lead
Nickel
Chromium (Hex)
Arsenic
Selenium
Manganese
Zinc
Cyanide
Chloride
Ammonia N
Sulfate
Conductivity (umhos/cm)
pH (S.U.)
Site Seepage Lime-Neutralized Sludge
0
0:
< 0
0
0
0
1
< 0
< 0
0
6
1
< 0
4,000
200
5,000
12,000
6
.3
.'01
.005
.04
.1
.5
.0
.01
.001
.005
.0
.0
.005




.5
0.2
0.01
< 0.005
0.04
0.2
0.03
0.1
< 0.01
< 0.001
0.005
0.02
0.05
< 0.005
50 2,000
5 10
1,200
1,600
9.0 9.4
Note: All results are in mg/1. Leachate results are based on an average or range of analyses of ex-
tractions from RCRA EP Toxicity Tests.
                                                                                                       POST CLOSURE
                                                           357

-------
testing of rosk was conducted at different intervals in open bore-
holes during drilling. Twenty-three test pits were excavated to eval-
uate soil conditions for  a borrow material. Shallow monitoring
wells were installed in the selected test pits.
   A total of 45 streams and seeps around the site area were sam-
pled. Private drinking water supplies within 1 mile of the site were
sampled and analyzed for sludge indicators. A total of 21 resi-
dences  were  identified, and monitoring wells  were sampled and
analyzed for chemicals indicative of the alkaline sludge leachate
from the existing disposal site: pH, total dissolved solids, chloride,
ammonia-nitrogen and nitrate-nitrogen.
   All sampling point locations are shown in Figure 3.


GEOLOGY

   The rocks that are exposed in the vicinity of the site belong to
two geologic groups. The oldest rocks, the Allegheny Group,  are
overlain by the Conemaugh Group,  both of Pennsylvanian age.
The top of the Upper Freeport Coal is the boundary between  the
Allegheny  and the Conemaugh Groups. The  rock  units are cov-
ered by unconsolidated material consisting of mine spoil and glacial
deposits.
   The entire area of the existing site was covered by glacial  de-
posits. This deposit occurs alike on hills or valleys as a mantle of
silt, clay, sand and boulders.  It is distinguished by a heterogen-
eous arrangement of this material, although locally it exhibits  the
effects of sorting by water action. In  most of the site area, strip
mining  for the Upper Freeport and Mahoning Coals has disturbed
the mantle of glacial deposits.
   A local valley occurs adjacent to the site. It is partially filled with
glacial deposits and stream sediments to a depth of 5 to 13 ft. The
maximum depth of till material is about 50 ft below the ground sur-
face, and its thickness averages from 20 to 30 ft.
   The Conemaugh and Allegheny Groups consist  of alternating
layers of sandstone, siltstone, claystone, limestone  and coal. The
Conemaugh Group  is represented at the site  by  the  Glenshaw
Formation. The older Freeport, Kittanning and Clarion Formation
represent the Allegheny Group (Fig. 4).
   The uppermost  rock unit present is the Glenshaw Formation.
It has been eroded or removed by strip mining over most of the site.
The upper beds at the site are composed of shale and sandy shale
which locally becomes sandstone. Below these  beds occurs a thin
coal seam  (Brush Creek Coal) which  is about  65  ft above  the
Mahoning Coal. Between these coal seams occurs a shaley unit with
a thick  layer of claystone named "The New Galilee Clay Shale."
The average thickness of this claystone is 6 ft. It has a characteristic
rusty color of orange brown or ochre. This claystone lies from 30
to 43 ft above the Mahoning Coal. A 6-ft thick shale layer sep-
arates the  Mahoning Sandstone from  the  Upper Freeport Coal
(Fig. 5).
   The rock unit between the base of the Mahoning Sandstone and
the top of  upper Kittanning Coal belongs to the Freeport Forma-
tion. The Freeport Formation and the upper portion of the Kit-
tanning Formation  included the Middle Kittanning Coal and its
associated underclay which were extensively studied. The top of the
Freeport Formation is  defined by the Upper Freeport Coal. The
seam has been removed  by both deep and strip mining. Where
present, it is approximately 3 ft  thick. The Freeport rocks and  the
upper unit of the Kittanning Formation is approximately  140 ft
thick. Below the Upper Freeport Coal occurs a continuous fossil-
iferous claystone and grey limestone. These rocks are in turn under-
lain by  a dark shale and. siltstone with sandstone layers and  the
Lower Freeport Coal.  The interval  between the Upper Freeport
Coal and the underlying Lower Freeport Coal, which is predom-
inantly shaley, is known as the Butler Sandstone. The thickness of
the Butler Sandstone ranges from 50 to 60 ft at the site.
  The Lower Freeport claystone (underclay) immediately below  the
Lower Freeport Coal is about 3 ft thick and grey in color.  Below
this strata  and between it and the  next  underlying  Middle Kit-
tanning Coal is an interval primarily  filled with dark and black
                                                                                    • TEST BORINGS
                                                                                    A STREAM OR SEEP SAMPLE
                                                                                  Figure 3
                                                                     Test Borings & Sampling Point Locations

1230-
I2OC-


MOO-

^
3
A
I
u IOOO-

9
$
z
5
5
u 900-
C


600-

73O-
*
I
i
^
Ł
5
*








t
§
^.
Ł
n

<



i
g
*
i



g
I
Q
a

If



	

|



l*&
•>/
*/

•*W

	

t T*?^|


•S"^"^"9



•••MHHHi


UH^^^^^^f


^•••••••H



C BRUSH CREEK COAL
NEW GAtXEC CLAYSTWC

I MAHONMS COAL (PARTIALLY DEEP MINEOI
\ CLAYSTDNEIUNOERCLAY)
• MAHONING SANDSTONE'
UPPER FREEPORT COM. (DEEP MMED)
^ CLAYSTONE (UNOERCLAY)
•BUTLER SANDSTONE*
C LOWER FREEPORT COAL
\ CLAYSTONE (UNOERCLAY)
•FREEPORT SANDSTONE"
"UPPER WORTHNGTON SANDSTONE"
^ MIDDLE KITTANNING COAL
\ CLAYSTONE (UNDERCLAY)
' LOWER WORTHINGTON SANDSTONE"
t LOWER KITTANNING COAL
\ CLAYSTONE (UNOERCLAY)

C CLARION COAL

                                                                                  Figure 4
                                                                          General Slratigraphic Section
358
POST CLOSURE

-------
                              WASTE  SLUDGE
                  .'.'y-UPPEH.-F.REEpORT-COAL'-'MINE VOIDS •, •. • '• •':' .• '
         ^*^- ••'•'••.••'.'•'; '•'•ByTLE'R'.SANDSfQNE •'." - '::'.' '.'••'. .'.'•'-'. '.' •'•.•'•'••!•;•".'••'.•
         '.-''.',•.•, . ^-'LOWER.'FREEPORT.-COAL'.' •'•-'•-'•  '. •'•••• -" •  • •' • ••-.'.-
         '.•"•.'.'.•'.'•'•'•' • '.'' .'••'.>."5EP°J
-------
                                       WASTE TREATMENT
                                                 ITIES
                           Figure 7
   Water Table Contours for Upper Freeport Coal Water-Bearing Zone
and Lower Freeport Coal. Both coals are separated by a confin-
ing layer of claystone. This layer has been eroded southward from
the site in the area of an unnamed tributary of Stateline Creek
and replaced by glacial deposits. The gap in the claystone aquitard
is of crucial importance inasmuch as it provides direct access for
some volume of recharge from the shallow aquifer to the under-
lain Butler Sandstone aquifer. Any leachate within  the shallow
aquifer can migrate directly into the deeper unit upon recharging
the window. In this area, groundwater is affected by leachate.
  Based  on the water level observations, this unit is  a semi-con-
fined aquifer.  The groundwater flow  is south west ward. This
aquifer is in the sandstone and fractured shale and siltstone. Lower
Freeport Coal also conducts the water flow. The hydraulic conduc-
tivity of this aquifer varies from 1.3x 10"4tol.7x 10"'cm/sec.
  The fifth principal aquifer in this area is the Freeport Sandstone
and Upper Worthington Sandstone above the Middle Kittanning
Coal. The continuous confining bed of the claystone creates a hy-
draulic barrier between the overlying Butler Sandstone aquifer and
the Freeport water-bearing  zone. The thickness of the confining
stratum varies from 2 to 12 ft. The groundwater flow direction in
the rock strata and Middle Kittanning Coal is westward and reflects
exactly the structure of Middle Kittanning Underclay.  In general,
this water-bearing stratum represents an unconfined aquifer. The
water table is only under a little pressure in the area  of the local
depression structure of Middle Kittanning Coal. The hydraulic con-
ductivities of this aquifer are  lower than in the  Butler Sandstone
aquifer and vary from 8.0 x 10"5 to 2.3 x 10"' cm/sec. The clay-
stone underlying the Middle Kittanning Coal is the confining bed
for this aquifer.

CONTAMINATION AND CONTAMINANT
MOVEMENT

  Five water-bearing zones were sampled for sludge leachate indi-
cators. Estimates of aquifer discharges and sludge  impoundment
water balances were performed. Chemical contaminants were con-
                                                       sidered to be soluble inorganic species, and conservative transport
                                                       in groundwater was assumed.
                                                       Shallow Unconsolidated Aquifer

                                                         This water-bearing zone consists primarily of mine spoils, alluv-
                                                       ium and glacial deposits below the sludge impoundment. Analyses
                                                       indicate that the aquifer is contaminated by leachate in the areas
                                                       directly south and west of the site. Chloride concentrations range
                                                       from  13.5 to 3406  mg/1  and average approximately 1200 mg/l.
                                                       No isoconcentration trends are apparent,  although a decreasing
                                                       concentration from  the site perimeter is expected. Total nitrogen
                                                       concentrations range from less than 0.1 to 52.75 mg/1. Average
                                                       total nitrogen is approximately 12 mg/1. Moderate concentrations
                                                       (200 to 500 mg/1 of chlorides) were found in the alluvium adjacent
                                                       to Stateline Creek, approximately 2500 to 3500 ft downstream of
                                                       the major deep mine discharge into Stateline Creek.
                                                         This area is probably affected by seepage from the creek bed into
                                                       the unconsolidated  material.  Stateline Creek  chloride concentra-
                                                       tions in this area are slightly higher than the range of chloride in
                                                       the alluvial test pit wells.
                                                       Mahoning Coal Aquifer
                                                         Analyses indicate that  Wells 31B-1 and  34B are contaminated
                                                       by sludge leachate.  Well  3IB-1 has chloride  levels from  115 tol
                                                       965 mg/1. Total nitrogen concentration is 28.4  to 43.6 mg/1 in
                                                       Well 31 B-1 and 0.47 to 19.9 mg/1 in Well 34B.
                                                         Wells 9-1, 28B and 33 all have chloride levels less than 25 mg/1
                                                       and total nitrogen levels less than 1 mg/1.  These wells  are in hy-
                                                       draulicaUy upgradient locations from the site.  Wells 31B-1 and 34
                                                       are also upgradient of the site; however, their chemical analyses
                                                       do not reflect their location. Well 34 is 20 ft from the edge of the
                                                       sludge pond, and although it is "upgradient," it is probably direct-
                                                       ly influenced by drainage from  sludge  deposits  at higher eleva-
                                                       tions and dispersion of contaminants from the sludge.
                                                         Well 31B-1 is apparently contaminated by the sludge leachates;
                                                       however, a hydraulic connection between the  well and the sludge
                                                       pond has not been confirmed. Well 31B-1  alone  might not be
                                                       sufficient to characterize the Mahoning Coal aquifer in the approx-
                                                       imately 60 acre area northwest of the site.

                                                       Upper Freeport Coal Aquifer
                                                         Analyses indicate that the deep mine pool on the northwest edge
                                                       of the site is contaminated  by sludge  leachate.  Well 3IB-2 has
                                                       chloride concentrations of 1200 to  1700 mg/1 and  total nitrogen
                                                       concentrations of 29 to 46.1 mg/1. Upgradient wells 8, 28C-1 and
                                                       32 had chloride levels less than 21.5  mg/1 and  total nitrogen levels
                                                       less than 1.0 mg/1.
                                                         Isochloride plots indicate that concentrations decrease with dis-
                                                       tance from the site as groundwater flows to the west. Well 35A-1,
                                                       approximately 1000 ft northwest of the site, had chloride levels of
                                                       422 and 870 mg/1.
                                                       Butler Sandstone Aquifer
                                                         Upgradient water quality, as determined from  water  table con-
                                                       tours, is  represented by Wells  1A-1, 2-1  and 28C-2. Chloride and
                                                       total nitrogen concentrations in these wells are less than 6.0 and
                                                       6.53 mg/1, respectively.
                                                         Analytical data for Wells 4-2, 5B-2,  27-2  and 35A-2 indicate
                                                       that significant contamination of the Butler Sandstone is occurring.
                                                       Chloride concentrations in these wells range from  100 to 1500 mg/1
                                                       and average approximately 965 mg/1. Total  nitrogen concentra-
                                                       tions in these wells range from 1.66 to 14.6 mg/1.
                                                         Isocontours for chloride indicate that contaminant migration is
                                                       to the northwest and southwest. No data are available beyond the
                                                       E. Fork  of Stateline Creek;  however, projections of the chloride
                                                       isocontours  indicate that concentrations greater than 200 mg/1 are
                                                       not expected west of the  creek. Increasing chloride concentration
                                                       trends are present for Wells 4-2, 5B-2 and 27-2. The cause of these
                                                       trends is not clear based on the available data.  The remaining wells
                                                       exhibit no significant concentration trends.
360
POST CLOSURE

-------
  Since the Stateline Creek drainage has eroded the Upper Butler
Sandstone in the site area, discharge of Butler Sandstone ground-
water into the Stateline Creek basin is expected.
Freeport Sandstone Aquifer

  Upgradient water quality as determined from water table con-
tours is represented by Wells 1A-2, 2-2, 3A-2, 28C-3  and 29A-3.
Chloride  and total nitrogen concentrations in these wells are less
than 95 and 7.9 mg/1, respectively.
  Analytical data for Wells 4-3,  5A-2, 5A-3, 6-3, 27-3  and 30-1
indicate that moderate contamination of the Freeport Sandstone
is occurring. Chloride concentrations in these wells range from 2.6
to 704 mg/1 and average approximately 244 mg/1. Total nitrogen
concentrations in these wells range from Co.l to 7.8 mg/1.
  Isocontours for chloride do not indicate any discernible contam-
inant migration trends. Comparisons of data from Wells 5A-2 and
30-1 exemplify the inconsistencies in isocontour trends. Wells 30-1,
which is approximately 1000 ft downgradient from Well 5A-3, has
a higher chloride concentration than 5A-3.
Seepage

  Seepage  was sampled at 32 locations.  Seeps were identified
around the perimeter of the facility, mostly on the slopes and at
the toe of the sludge impoundment dikes.
  Chemical analyses  indicate that  sludge impoundment seepage
has the highest concentrations of leachate indicator parameters of
any samples taken in the site area.  Chloride concentrations range
from 894 to 12,680 mg/I. Average chloride level is approximately
5650 mg/1.  Total nitrogen concentrations at these points range
from less than 1 mg/1 to 460 mg/1.
   Embankment seepage and  shallow groundwater were found to
have the highest concentrations of leachate indicators of all waters
analyzed at the site.  Upper Freeport Coal deep mine discharges
had significantly lower concentrations of indicators, probably be-
cause of dilution  from upgradient drainage within the deep mine.
The two aquifers below the Upper Freeport Coal, the Butler and
Freeport Sandstones, have minor concentrations of sludge leach-
ate indicators.
   The major groundwater discharge at the site is the Upper Free-
port deep mine drainage. Discharge points near Stateline Creek
had an average combined flow of 200 gal/min for  approximately
2 yr of recording in 1977 and  1978. These flows have been relative-
ly constant over the last 10 yr of observation.
   The presence of the extensive deep mine workings in the site area
is a positive factor in reducing leakage to the Butler and Freeport
Sandstones. The deep mine  openings provide an  effective free-
draining zone that acts as a horizontal groundwater  interceptor.
Leakage from the  Mahoning Sandstone above and  from the
shallow groundwater system is drained through the mine and dis-
charged. This minimizes the standing water in the deep mine and
reduces leakage into the lower Butler Sandstone.
   Seepage flowrate from the Mahoning Coal and Sandstone and
the  sludge pond  embankments is significantly less  than the deep
mine discharge. These discharges have not been accurately meas-
ured, but visual estimates of total average flow range from 20 to
50 gal/min.
   Estimates were made of the discharge rates  of the Butler and
Freeport Sandstones into the drainage basin of the  E. Fork of
Stateline Creek using Darcy's Law. The transmissive area was cal-
culated by multiplying the aquifer  saturated  thickness by the dis-
charge length. The discharge length into the creek is estimated by
the  distance between croplines of the aquifer  confining units,
namely the coal outcrops.
   The Butler Sandstone average discharge  rate is estimated at
36.4 gal/min. The Freeport  Sandstone average discharge  rate is
estimated at 0.72 gal/min.
  A water  balance  was estimated  for the sludge impoundment,
both prior to and after final closure. The water balance calcula-
tion estimates the amount of liquid inflow to the sludge, which on
a steady-state basis equals the liquid outflow, or  leachate flow,
from the sludge. An annual budget was calculated.
  The water balance evaluated the sources of moisture inflow to
the sludge. These flows include the lateral groundwater flow from
the highwall above the Mahoning Coal and direct precipitation on
the sludge surface prior to closure.
  A significant difference was estimated for sludge pond inflow for
the pre-closure and post-closure  periods. Annual  average  pre-
closure inflow is  43 gal/min, and post-closure is 2.37 min.  The
primary contributor to pre-closure inflow is direct percolation on
to the open sludge surface which accounts for an average of 40.6
gal/min,  or 95% of the estimated  annual inflow. After topsealing
of the site, direct percolation is assumed to be zero,  and the total
annual inflow contribution will be from groundwater. The lateral
groundwater inflow contributes 1.25 million gal/yr, or 2.37  gal/
min of inflow to the sludge from the Mahoning Coal highwall.
  Groundwater inflow is assumed to be present during the pre- and
post-closure periods, since topsealing of the sludge pond is not ex-
pected to affect the groundwater above the Mahoning Coal.
  A summary of the groundwater discharge rates and impound-
ment water balance is contained on Table 3.
REMEDIAL CLOSURE DESIGN
  Site investigation findings determined the extent of contamina-
tion and  the characteristics of the subsurface drainage  system.
The objectives of the remedial cleanup were to:
•Reduce  contaminant loading on  Stateline Creek to acceptable
 State and Federal discharge levels
•Prevent groundwater contamination of the Freeport Sandstone
 and lower aquifers
  RCRA requirements for closure, post-closure and groundwater
monitoring plans  were addressed in development of the remedial
closure design. These requirements are  outlined in  the  PADER
regulations for hazardous waste management (25 PA Code Chapter
75, Nov. 29,1980).
  The remedial concept determined to be  the most cost-effective
was capping the sludge impoundment and collection and treatment
of leachate and contaminated groundwater discharges to Stateline
Creek.
  A conceptual closure plan is depicted on Figure 8. The sludge
impoundment discharge would be significantly reduced by placing
a low permeability topseal on the impoundment surface. Leachate
drainage from the impoundment would be  collected by a series of
interceptor trenches located at the Upper Freeport Coal deep mine
drainage, in the shallow unconsolidated aquifer and  at the Butler
Sandstone  aquifer discharge zones.  Collected wastes would be
treated on-site to  acceptable discharge limits and then discharged
into Stateline Creek. The design of the wastewater treatment facil-
ity is not discussed in this paper.
  PADER agreed to  permit limited leakage of slightly  contam-
inated groundwater from  the Butler Sandstone aquifer  into the
                           Table 3
         Aquifer Discharge and Impoundment Water Balance
Source
Discharge
(gal/min)
                                    Receiving Body
  Mahoning Coal Aquifer

  Sludge  Impoundment
  Pre-Closure
  Post-Closure
  Upper Freeport Coal Aquifer

  Butler Sandstone Aquifer


  Freeport Sandstone Aquifer

  Shallow Unconsolidated Aquifer
     2.4  Sludge Impoundment


    43    Upper Freeport Coal Aquifer
     2.4  and Shallow Unconsolidated
          Aquifer

   200    Stateline Creek

    36    Shallow Unconsolidated
          Aquifer and Stateline Creek

     0.7  Stateline Creek

   20-40  Stateline Creek
                                                                                                     POST CLOSURE
                                                          361

-------
                                       WASTE TREATMENT
                                       ?ACILITIES
                                                                                                          «" »u«'Ace LA»C« SHALL BE
                                                                                                                          TUftC
                                                                                           lUfffH t tUft'ACC LATCR SHALL Bl
                                                                                           I rOPIOIl. NATUHAL SOIL. 00 A Mill
                                                                                          floe »OTM
                                             RUNOFF DIVERSION/
                                             COLLECTION
                                 [XI TREATMENT PLANT
                            Figure 8
                     Conceptual Closure Plan
 Freeport Sandstone aquifer. The Freeport Aquifer was not signif-
 icantly degraded, and implementation of remedial measures is ex-
 pected to improve drainage quality into the aquifer. Post-closure
 monitoring of the Freeport aquifer was required to assess ground-
 water quality changes.
   Field studies were undertaken to evaluate final cover designs for
 the impoundment topseal. PADER required a cap with a hydraulic
 conductivity of less than 1 x 10"' cm/sec.
   A  1 to 2 ft layer of solidified sludge from the process pilot plant
 already present on the impoundment surface was evaluated as a
 topseal material. Nine test cells of solidified sludge, 12 ft x 12 ft x
 2 ft thick, were  constructed on the impoundment surface. Exist-
 ing placed sludge and various fresh mixes of processed wastes were
 tested.
   The results  of the field testing indicated permeabilities of the
 solidified sludge  for all test cells averaged 2 x 10"' cm/sec ranged
 from 3.9 x 10~»  to 9.8 x 10"' cm/sec. Thus, it did not meet the
 permeability criteria and therefore was not acceptable as a topseal.
   An alternate design had to be used  (Fig. 8). A 20 mil PVC non-
 reinforced membrane was selected. It was protected on both sides
 by a non-woven geotextile fabric.
                                                                           VCOCTATION «*St flLL



                                                                        10'luin) GRAOC COABSC FILL
                                                                                              OCOTOTILC
                                                                                              to mu rvc MCUIIIANC o* A»MOVCO
                                                                                              (OUAL
                                                                                               :OTl«TILt 'ABDIC
                                                                                    Figure 9
                                                                               Sludge Pond Topscal

                                                            At anticipated cover surcharge thicknesses of 6 to 10 ft, primary
                                                         (90%) settlements of 5 to 7 ft over 150 to 280 days were estimated.
                                                         These equate to approximately a 10% settlement.
                                                            Since the disposal facility was approximately 23 yr old, most of
                                                         the sludges were expected  to  achieve 90% or greater settlement.
                                                         However, since significant  volumes of sludge were disposed in the
                                                         last 2 to 3 yr of operation, surcharging of the impoundment and
                                                         monitoring of settlements was proposed as a first stage of closure.
                                                            In addition to placement of low permeability topseal and con-
                                                         struction of a leachate collection  and treatment system, extensive
                                                         grading, re vegetal ion and surface drainage measures were imple-
                                                         mented to minimize rainfall recharge into the local groundwater
                                                         system and reduce long-term site erosion.
                                                            A 5  acre impoundment on the  main watercourse  below the im-
                                                         poundment was used as the sedimentation basin for closure plan
                                                         construction. This upper pond is topographically situated and has
                                                         sufficient capacity to provide the minimum storage requirements
                                                         for stormwater sedimentation. The pond  and two lower ponds
                                                         along the watercourse were upgraded with stable overflow  spill-
                                                         ways to protect against excessive erosion.
                                                            Stage 1 closure operations commenced in June,  1983. A mini-
                                                         mum 2-ft thick cover of local mine spoils was placed over the entire
                                                         impoundment. Two 10-ft thick surcharge fills were constructed on
                                                         the two longest impoundment areas  in the central pond. Settle-
                                                         ment plates were installed on  the surcharge fills and surveyed for
                                                         elevation and horizontal control. Approximately 220,000 yd* of fill
                                                         were placed in Stage 1.
                                                            Stage 2  construction commenced in May,  1984.  Work in this
                                                         stage includes regrading of the cover fill, placement of the cover
                                                         topseal, construction of the surface drainage and leachate collec-
                                                         tion systems and  placement of the  final  cover and vegetation.
                                                         Approximately 300,000 yd' of fill will be placed in  Stage 2. Stage
                                                         2  is presently under construction  and is scheduled for completion
                                                         in October, 1984.
                                                            Special considerations were made in construction  of the topseal
                                                         for the 45 acre impoundment. The settlement potential of up to 70
                                                         ft  of semi-solid sludge had to be  accounted for to prevent future
                                                         slumping of the final cover surface and ponding of rainfall water.
                                                            Calculations were  made to estimate  primary and  secondary
                                                         settlements of the sludges. Sludge properties were taken from lab-
                                                         oratory consolidation tests. Basic assumptions included a homog-
                                                         eneous and i so tropic medium, instantaneous placement and one-
                                                         dimensional vertical consolidation.
362
POST CLOSURE

-------
                                   BENEFICIAL REUSES  OF
               HAZARDOUS  WASTE SITES  IN  CALIFORNIA

                                             JULIE K. ANDERSON
                                    U.S.  Environmental Protection Agency
                                     Toxics & Waste Management Division
                                            San  Francisco,  California
                                          HOWARD K. HATAYAMA
                                   California Department  of Health Services
                                                Berkeley, California
 INTRODUCTION

  The California Department of Health Services (DOHS), Toxic
 Substances Control Division, is the governmental body with pri-
 mary responsibility for the management of hazardous wastes in
 California. Laws governing hazardous waste management were in-
 itially enacted in California in 1972. Since that time the state has
 developed and implemented a variety of related programs dealing
 with facility permitting, regulation of generators and transporters,
 development of alternative disposal technologies, site mitigation
 and management of the "State Superfund."
  In 1979, the Department of Health Services initiated its "Aban-
 doned Site Project", designed to seek out properties throughout
 the state where  hazardous wastes were once stored, disposed of
 or otherwise handled. Since its inception, dozens of such sites have
 been identified,  many of which still contain hazardous wastes in
 piles, lagoons, pits, sumps, storage tanks and/or drums. In many
 cases, these wastes have leaked from their containers, contam-
 inating adjacent properties and underlying waters. The persons
 originally responsible for such contamination  have frequently
 died, filed bankruptcy, relocated or sold the parcels to land devel-
 opers desirous of  utilizing the properties for purposes often en-
 tirely unrelated to their former uses.
  The Department of Health Services is particularly interested in
 sites slated to undergo redevelopment. While strongly in favor of
 restoring these former  hazardous waste sites  for beneficial new
 uses, DOHS also takes a very cautious approach to redevelopment,
 carefully considering the potentially severe long range effects such
 reuse can have  if the future intended  use is incompatible with
 wastes still remaining on-site. DOHS therefore works closely with
 these  developers, taking an active role in the characterization of
 site contamination and  in the design of remedial measures which
 will be fully protective of the health and safety of future site occu-
 pants.
  Mitigation strategies designed for redeveloped hazardous waste
 sites must be  individually tailored  to each situation, taking into
 consideration  a  number of vital decision factors. In  this paper,
 the authors present the case histories of three such sites  in Cal-
 ifornia,  focusing on  the cause and nature of the contaminants
 present, the proposed reuse of each site, selected mitigation strat-
 egies and (most importantly) the various decision factors consid-
 ered in the selection of these mitigation strategies.

BOUCHER LANDFILL
  The Mola Development Site, formerly known as the Boucher
Landfill, is located in the city of Huntington Beach, California on
a low plateau overlooking a wildlife refuge. The 12.4 acre site was
originally used as a gravel pit in about 1934. In the 1940s, the site
was enlarged and a variety of petroleum refinery-type wastes, in-
cluding acidified sludges from the production of high octane gas-
oline, were deposited  in the excavation.  In the early 1950s, a
permit was granted to deposit rotary drilling muds at the site.
  It was shortly after this period that adjacent residents began  to
complain about petroleum or gasoline odors and taste in their well
water. In 1963, the site was permitted as a Class II  site allowing
disposal of building demolition wastes. By 1970, more than 12 pub-
lic and private wells were closed in the area adjacent to the site due
to contamination by materials deposited at the site.
  In 1979,  the new  owner of the property,  Mola Development
Corp., applied to the city for a conditional use permit and tenta-
tive tract map to construct a 224 unit residential development on
the site. After deliberating on six separate occasions on the appli-
cation, the city planning commission decided to pursue a negative
declaration  for the project on the conditions that additional in-
formation be submitted on the nature and extent of contamina-
tion and that several mitigation operations be developed for the
site.
Extent of Contamination

  The subsequent site investigation consisted of 20 boreholes rang-
ing in depth from  15 to 25 ft. Nine of these boreholes were con-
verted to gas/vapor sampling wells. Sixteen existing groundwater
wells were located nearby. Five of these were selected for samp-
ling based on their  proximity to the site and their completion in the
shallow aquifer. The investigation revealed large pools of black,
viscous and extremely odorous petroleum wastes at various depths
in the fill. At a few locations, this material was found to be oozing
to the surface. Due to the porous nature of the fill, surface water
had percolated through to leach many constituents of the waste
material into surrounding soils and groundwater. A 1963 estimate
of the amount of sludge deposited in the fill was 2000 yd3. The re-
vised estimate as a  result of these borings was 60,000 yd3 of sludge
and contaminated fill. The constituents of primary concern in  this
material were aromatic hydrocarbons (benzene and toluene), chlor-
inated hydrocarbons, phenols and a host of organic sulfur com-
pounds (thiophenes). In addition, the waste exhibited very low pH
(pH 1-2).
  Many of the borings bottomed in sandy lenses indicating some
potential for off-site migration. Well sampling showed that some
waste constituents, particularly benzene, had migrated almost a
mile off-site in the shallow zone.
                                                                                               POST CLOSURE
                                                      363

-------
  The landfill gas contained  high concentrations of benzene  and
other aromatic and  aliphatic  hydrocarbons, including  methane.
Thiophenes  were also found  in abundance. The calculated  gas
emission rate  based  on the gas  well readings  was approximately
600ft'/day.

Mitigation Strategies and Decision-making Factors

  Several alternatives were evaluated for effectiveness and feasibil-
ity of controlling soil contamination, vapor generation, vapor mi-
gration,  odors and leachate migration.  These  alternatives ranged
from no action/no project to  in situ treatment, encapsulation with
gas control and  excavation. Despite the findings of the  investiga-
tion, the developer was still interested  in pursuing the project. Be-
cause of the uncertainties related to the effectiveness of the in  situ
treatment and encapsulation alternatives, the City Planning Com-
mission decided  to accept the excavation alternative recommended
by the regulatory agencies. Concerns about the potential for migra-
tion of  toxic and odorous compounds into residences built on the
Till after treatment or encapsulation, about the furlhcr contamina-
tion of the aquifer and about the potential for  liquefaction in a
seismic event were key factors in this decision.
  Citizen concerns about  the potential health  and environmental
impacts  of the project prompted the Commission to order a  full
Environmental Impact Report (EIR) for the project which involved
removal of an  estimated  105,000  yd'  of material  at a cost of
approximately $2,000,000. The EIR identified  the need for a very
controlled excavation process, ambient air monitoring, a specified
transportation route, special  handling at the disposal site and a
community evacuation plan. These special measures were required
primarily because of the gases and vapors in the waste which would
be  released upon excavation.  Despite these further constraints on
the project, the developer was still intent on proceeding. The EIR
and the project were finally approved in mid-1981.
  The removal activity took  place  over a period of 2-3 months;
this period included several work stoppages as a result  of exces-
sive odor complaints from  the community.  The estimated  volume
of  material excavated exceeded the initial estimates by over 50%
and contamination was found to a depth of 30 ft in some portions
of the site. The excavation phase was completed in July of 1981.
  Building permits  were finally issued in  June  1983. Currently
there are 288  condominium-style residential units on the  former
hazardous waste site. The units include studios, one bedrooms  and
two bedrooms with prices ranging from $69,000 to $130,000.
  This  is one of the first  uncontrolled hazardous waste  sites in
California where redevelopment has occurred and where land use
has been upgraded to such an extent. The process involved a care-
ful evaluation of the potential health and environmental  risks in
allowing residential use of a site formerly used for disposal of haz-
ardous wastes. At the time the decision was  made, there was little
confidence in  the effectiveness of any measure other than  excava-
tion in  protecting the health of residents of the condominiums.
For this type of site, the same decision would probably be made
today. Not enough is known about the effectiveness of in situ con-
trol measures for migration of gases and leachate  to ensure  that ex-
posure of residents will not occur in  such  high-risk situations as
redevelopment of contaminated lands for residential purposes.
  The incentive for  redevelopment  of contaminated lands to  uses
which  would  provide the highest potential  for exposure  to con-
taminants is clearly tied to  the high potential for return on invest-
ments for residential uses. The desirable locale of a site  (one mile
from the beach, in this case) is a key factor in determining whether
remedial  action  to the extent  that residences can be built is eco-
nomically feasible. This type  of redevelopment  is more econom-
ically feasible in urban areas such as San Francisco and Los Angeles,
where the demand for residential housing is high.


BETHLEHEM STEKI. COMPANY

  South San  Francisco,  nicknamed "The  Industrial  City",  was
founded in the  1800s as an inexpensive alternative  for  industries
                           Figure 1
   Condominiums at Mola Development Site, Huntington Beach, CA.
                  Former Site of Boucher Landfill
seeking the advantages of San Francisco's port location. Numer-
ous businesses, notably the meat packing, paint manufacturing and
steel industries, established their operations there at the turn of the
century.
  The 110-acre site now owned by Homart  Development Com-
pany was occupied by steel manufacturing companies from  1903
to 1977.  Bethlehem Steel operated a full scale steel  production
facility there complete with open hearth blast furnaces,  a milling
and annealing operation, welding shops and a galvanizing facility.
In 1977, Bethlehem Steel vacated this property, in part due to the
heavy costs of complying with environmental regulations.
  The site  was then  purchased  by Homart  Development Com-
pany with  the intention of constructing the  multi-million dollar
"Gateway  Center" project,  a  combination  hotel/commercial/
office park development. In 1980, this as yet  undeveloped prop-
erty came to the attention of California's Abandoned Site Project
as a  likely location  of hazardous wastes, and characterization
efforts were cooperatively initiated by Homart and the Department
of Health Services.

Extent of Contamination
  Throughout Bethlehem Steel's long history of operation, slag
from the steel manufacturing process was disposed of at various
locations on the  property. These  widespread slag deposits  con-
tained hazardous concentrations of heavy metals; particularly lead,
zinc  and chromium.  In addition,  the open hearth building and
surrounding property  were heavily contaminated with metals from
furnace ash deposition.
  PCBs were also detected throughout the site at electrical shops,
transformer storage areas and anywhere heavy machinery had  been
employed. An underground storage tank containing fuel oil for the
hearth and mill operations also  contained PCBs, but in concen-
trations less than SO ppm. An investigation of this area determined
that  substantial leakage from the tank had occurred, resulting in
a subsurface plume of oil approximately 120  by 500 ft. Though
the resulting contaminated soils contained only low concentrations
of PCBs, some free liquids were still present  in underlying  frac-
tured bed rock.
  Investigation of the  underlying groundwater detected only minor
concentrations of dissolved metals  confined to areas  formerly
used as acid seepage basins for disposal of pickling liquids. How-
ever, these  low concentrations were deemed an insignificant haz-
ard to human  health or the environment; the  high salinity of the
          POST CLOSURE

-------
                           Figure 2
  Homart Development Company's "Gateway Center", Site of Former
           Bethlehem Steel Plant, So. San Francisco, CA
•The one-foot deep cover must be maintained at all times.
•Excavation will be  allowed below the cover provided DOHS is
 informed in advance and any excavated soils are relocated within
 property boundaries beneath the one-foot deep cover. Soils which
 cannot be relocated on-site must undergo laboratory analysis with
 subsequent disposition subject to DOHS approval.
•Precautions to prevent surface runoff and windblown dust must
 be undertaken whenever excavation occurs. Appropriate worker
 safety precautions must also be utilized during excavation.
•Future use of the site is limited to commercial,  light industrial,
 office park and hotel purposes. No residences, schools, day care
 centers or hospitals may be built.
•The deed restriction transfers  to all future tenants and subsequent
 landowners.
   The above mitigation strategy was approved by the regulatory
agencies for the following reasons:
•Upon investigation, groundwater contamination was deemed neg-
 ligible and relatively immobile and the underlying aquifer largely
 unusable due to high salinity.
•Contaminants remaining on-site were limited to relatively immo-
 bile, low toxicity metals.
•All future surface runoff, windblown dust and direct contact with
 hazardous materials remaining on-site will be precluded via main-
 tenance of the one-foot deep cover. This precaution will further
 be enhanced through landscaping and through paving of the site
 for roadways and parking lots.
•Redevelopment of the site will be limited to industrial/commer-
 cial uses only, thus reducing the likelihood of any inadvertent
 or long-term exposure to buried metallic wastes.
•Total removal of metallic wastes  was deemed unfeasible from
 both  the standpoints of economics and of limited local landfill
 disposal capacity due to the depth and extent of contamination
 on this 110-acre site.
   To date, Homart  Development  Corporation has fully  imple-
mented the control strategy required by DOHS. The site has now
been fully graded and partially landscaped. A partially occupied
12-story office structure has been erected overlooking San Fran-
cisco Bay. Roadways and a connecting ramp to the nearby free-
way have been constructed. It  is anticipated that this complex of
hotels, research/development facilities, office space and commer-
cial/retail buildings will be  completed by 1987, providing both an
economic boost and a facelift to this former industrial area.
aquifer precluded its use, and migration potential was determined
to be severely limited.

Mitigation Strategy and Decision-making Factors

  The following mitigation measures were undertaken by Homart
Development Company following approval by DOHS:
•All laboratory chemicals from the quality assurance laboratory
 and other drummed wastes were identified and properly disposed
 of.
•The underground oil  storage tank was  pumped,  steam-cleaned
 and the resulting liquids properly disposed of.
•All soils containing hazardous concentrations of PCB were ex-
 cavated and removed.
•Fractured  bedrock containing liquids contaminated  with PCB
 was trenched, dewatered and the resulting  effluent  separated
 into oil and water fractions. Remaining  oily  residues clinging to
 excavated materials were collected on absorbent material for dis-
 posal. Resulting nonhazardous solids were ultimately returned to
 the site for use as fill.
•Certain areas exhibiting low pHs were neutralized with  lime to
 further deter mobility of metals.
•The site was graded and covered with 1 ft of compacted clean soil.
  A restrictive covenant was placed upon  the deed to this site and
was entered  into public record in  the county of San Mateo. This
document, which is legally enforceable by DOHS and effective into
perpetuity, provides the following:
HERCULES POWDER COMPANY

  The small company town of Hercules, California grew up around
the Hercules Powder Company, an explosives manufacturer which
once owned approximately 1300 acres of land in the area. This
industrial plant, located in rolling hills 30 miles north of San Fran-
cisco, operated as a powder works from 1885 until 1976 and man-
ufactured such chemicals as methanol, ammonia,  formaldehyde
and nitric acid. These materials were,then transformed into fertil-
izers and explosives including trinitrotoluene (TNT), nitroglycerine
and black  powder. For safety purposes, the various explosives
production lines were scattered throughout the area and located
as much as a mile from the primary chemicals plant.
  In 1976, Hercules Powder Company sold its west coast plant to
Valley Nitrogen Producers, who continued to manufacture fertil-
izers there until 1979. At that time, the formerly large  holdings
of Hercules Powder Company were subdivided and sold in smaller
parcels to a variety of land developers  and holding companies.
These firms, working with the City of Hercules, developed a master
plan to transform Hercules from a one-industry town into an inte-
grated series of commercial and light industrial developments,
office parks and residential communities.
  The California Abandoned Site Project became aware of this
former explosives/fertilizer operation  in 1980. At that time, the
City of Hercules  approached DOHS with its area-wide  develop-
ment proposal and requested aid to resolve any hazardous waste
issues. Since then, DOHS has worked closely with many  recent
                                                                                                   POST CLOSURE
                                                         365

-------
purchasers of these parcels to characterize and mitigate contam-
inants present on their sites. Indeed, consideration of the potential
presence of hazardous wastes on these properties has become a
vital component of the Environmental  Impact Report  process for
the area.

Extent of Contamination

  Each  of the several parcels sampled  has  elevated concentra-
tions of heavy metals, notably lead and zinc. These metals appear
to be associated with spilled catalyst  materials or empty decay-
ing drums strewn  about  the premises.  Several areas also contain
persistent, toxic, organic explosives, such as trinitrotoluene (TNT),
dinitrotoluene (DMT) and dinitrobenzene (DNB). One area, used
primarily for the discharge of contaminated wastewaters, contained
one  lined and two unlined  discharge  ponds containing  metallic
sludges  and/or explosives. Another parcel  (the one most heavily
contaminated  with explosives) contained the remains of the old
TNT production facility along with two large, submerged urea
tanks. A third parcel housed the black powder line. Old  bunkers
contaminated  with metals and explosives were found at  several
sites, and many aboveground storage  tanks  (some still contain-
ing liquid chemicals  or residues) were  located at the old  primary
chemicals plant.
  Groundwater monitoring has detected no significant  contamina-
tion of the underlying aquifer which is  considered too  saline to be
used for most beneficial purposes. Residents and businesses in the
area draw their water only from the municipal water supply.

Mitigation Strategies

  So far,  formal mitigation  plans have  been  developed and
approved for five of the seven sites currently characterized.  These
plans are based on the nature of the contamination and the future
intended land use.
  Three of the sites with approved mitigation plans are to be de-
veloped for residential purposes  including condominiums, single-
family dwellings, schools, parks  and playgrounds. They were re-
quired to undertake stringent cleanup measures including total re-
moval  of contaminants  to DOHS-determined  nonhazardous  or
background concentrations.
  The other two sites with approved mitigation plans are to  be
used for commercial and light  industrial purposes. The owners of
these sites were offered the options of: (1) totally removing site
contaminants to nonhazardous levels, or (2) removing heavily con-
taminated soils to some intermediate cleanup level established  by
DOHS,  encapsulating remaining hazardous wastes on  site and ac-
cepting a deed restriction on those areas still containing hazardous
wastes.  One site owner opted to completely remove wastes, find-
ing this  alternative preferable to the perpetual restrictions imposed
by a deed covenant.  The other owner has,  for economic reasons,
tentatively chosen  to relocate all remaining wastes to a deep gulch
area on-site where encapsulation will occur. This area  will be sub-
ject  to a deed  restriction  similar to that currently in  effect for the
Homart Development site, while clean areas  of the property will
be available for unrestricted future use.
  Factors considered in choosing these mitigation strategies include
the following:
•Groundwater  contamination was nonexistent or negligible at  all
 sites. Underlying water was also of limited usefulness  due to high
 salinity, and all residences and businesses in the area are to be con-
 nected  to municipal water supplies.
•Stringent cleanup measures resulting in unrestricted future prop-
 erty uses were deemed essential at residential, school and play-
 ground properties for several reasons:
•Occupants of residential-type properties, and particularly
 children, would have a  higher likelihood of  contacting haz-
 ardous waste residues left in  place on these  properties than
 would  workers  employed  on commercial/industrial sites.
 Mechanisms  for  such  exposure would include  playing  of
 games  such as baseball on dusty, undeveloped lots  or  un-
                                                                                   Figure3
                                                           Citation Homes, One of Many Developments Under Construction at
                                                             the Site of the Old Hercules Powder Company, Hercules, CA.
                                                         paved schoolyards, ingestion of soils by small children play-
                                                         ing in unlandscapcd backyards, ordinary gardening and land-
                                                         scaping (followed by consumption of foods without prior
                                                         washing of hands), ingestion of foods  grown in contam-
                                                         inated gardens and possible installation of domestic wells
                                                         by persons lacking knowledge of groundwater use criteria.
                                                        •Residential occupants may represent a more vulnerable pop-
                                                         ulation  than workers at a commercial site, as small children,
                                                         the elderly and invalids are more likely to be represented in
                                                         the former group. In addition,  workers are protected to some
                                                         extent by occupational safety  organizations, whereas home-
                                                         owners are not.
                                                        •Residential occupants exposed to hazardous wastes at their
                                                         homes are more likely to experience a longer exposure period
                                                         than  workers  who spend  a finite amount of time at their
                                                         workplaces.
                                                        •Contaminants present included not only metals but relatively
                                                         persistent organic explosives.  While not an explosive haz-
                                                         ard under site conditions,  these compounds do exhibit mod-
                                                         erate to high toxicity via several routes of exposure. In addi-
                                                         tion,  DMT is a suspected carcinogen. Considering these tox-
                                                         icities and the fact that these materials are not normally en-
                                                         countered in the natural environment, extra levels  of precau-
                                                         tion were deemed prudent in residential scenarios.
                                                        •It was considered unlikely that  DOHS could effectively police
                                                         land use restrictions imposed  on private residences due to
                                                         the large number of properties  affected.
                                                        •Lowered property values may  result at residences  located on
                                                         hazardous waste property subject to deed restrictions.
                                                           For  all of the above reasons, complete removal of hazardous
                                                        wastes from residential properties  was considered  to be the miti-
366
POST CLOSURE

-------
gallon alternative most protective of public health and safety as
viewed by DOHS and the most desirable as viewed by the devel-
opers involved.
•Land use restrictions were  offered as an alternative mitigation
 to those developers intending to use their properties for commer-
 cial or industrial purposes for several reasons:
•The control  measures included in such restrictions  would
 effectively prevent  all surface runoff, windblown dust  and
 direct contact with underlying wastes.
•Any exposure  which may  inadvertently occur via excava-
 tion activity would be of limited duration  due to the in-
 tended use of the property. For the same reason, excava-
 tion activities would  also be subject  to control by occupa-
 tional safety organizations.
•All very hazardous contaminants, as determined by DOHS,
 would be removed from  the  property prior to encapsula-
 tion and imposition of the deed restriction.
•Widespread contamination could render complete removal of
 wastes unfeasible and undesirable, both economically  and
 from the perspective of available landfill capacity.
   Today,  cleanup and control  of the hazardous wastes from the
old Hercules Powder Plant is well underway. More than 12,000
yd3 of contaminated soils have been excavated, and the carefully
planned development of the  City of Hercules  has been launched.
Two model home developments  are now open.  When complete,
this region-wide project will preserve elements of this town's color-
ful past while updating the community with new housing facilities
and increased employment opportunities.
CONCLUSIONS
  Redevelopment of former hazardous waste sites can be success-
fully accomplished when certain precautions are taken to fully pro-
tect the health and safety of all future site occupants.
  Factors to be considered by regulatory agencies in designing mit-
igation strategies for such sites must include the nature of the con-
taminants present (i.e., their degrees of hazard via various routes
of exposure, mobilities in  soil systems,  persistence, degradation
potential, volatility,  etc.), the actual and potential extent  of con-
tamination in soils, air, surface and groundwaters, beneficial, uses
of these potentially effected water resources, the feasibility and re-
liability of all  available mitigation strategies and the intended
future land use.
  Land developers must consider  an  additional set  of decision
factors when confronted with agency-approved mitigation alterna-
tives. Among these factors  are the cost of each mitigation strategy
(in time and  money), potential resale value of redeveloped sites if
those sites still contain waste residues,  the need for possible long-
term involvement with a site via pollutant monitoring or deed re-
strictions, overall return on investment and potential  public rela-
tions issues.
  By working together cooperatively on all phases of site investiga-
tion and remedial  design, land owners, city development  author-
ities and environmental regulatory agencies can ensure  that former
hazardous waste sites may  be redeveloped for beneficial new pur-
poses while serving the community in productive capacities which
are also fully protective of public health and safety.
                                                                                                    POST CLOSURE
                                                         367

-------
       PUBLIC INFORMATION  NEEDS  IN  THE  SITING AND
                  CLEANUP OF  HAZARDOUS  WASTE  SITES

                                              NANCY J. JERRICK
                                                 CH2M Hill,  Inc.
                                                 Portland, Oregon
                                                NANCY R. TUOR
                                                 CH2M Hill,  Inc.
                                                  Reston,  Virginia
THE PUBLIC'S NEED TO KNOW
  Some basic tenets of public participation in  hazardous waste
activities have become clear from  the experience of the past few
years:
•The public will continue  to be interested and involved in issues
 surrounding the siting, expansion and cleanup of hazardous waste
 sites.
•The public has the right to  know  and participate  in  decisions
 affecting health and welfare. The public recognizes this right.
•Public involvement in the decision-making process is  absolutely
 essential in finding and implementing acceptable hazardous waste
 solutions.
  Government, industry and  the public have been struggling to
create effective public participation  processes that address these
basic facts. The Superfund program strongly emphasizes two-way
communication between communities and government throughout
all Superfund activities and has continuously refined its policies
and procedures in response to  what has been learned over the past
four years.  RCRA's community relations program is also being
expanded. Industry has learned—sometimes painfully—that under-
estimating the public's need to  know is done at its own peril.
  "Underestimating"  and "peril" are key words. The public has
become increasingly sophisticated about environmental and health-
related issues. Grass-roots politics, the environmental movement
and consumer information programs have provided skills and in-
sights to people who are demanding accountability from those
who affect their lives. Unfortunately, the generation and disposal
of hazardous  waste have  not always been entirely accountable,
intensifying the public's apprehension and its resolve. People arc
afraid. Uncontrolled hazardous waste is dangerous, and the danger
can be long-lived. People are educating themselves about how they
and their children are being affected and how they,  in  turn, can
have an effect.  This combination of intense concern,  increased
awareness and political know-how  has given the  public a powerful
voice in the decision-making process.

WHAT IS AN INFORMED PUBLIC?

  There is general  agreement that any future  siting,  operation
and cleanup of  hazardous waste facilities will  be a joint—and,
ideally, cooperative—effort among government, industry and the
public. Processes for public involvement continue to evolve.  The
goal of  these processes is  to  foster responsible, effective partic-
ipants—in short, an informed public.
  The question then arises of  the meaning of "informed." What
do people  need to know to  weigh the alternatives  and make
                                                     choices? What do they want to know when their lives are directly
                                                     touched by some aspect of hazardous waste? What kind of public
                                                     education programs and materials can provide this information?
                                                       Past experience has shown that a number of central issues are
                                                     commonly raised  by the public, and  that adequate information
                                                     has not always been provided to address them. These issues and
                                                     information needs are discussed below. First, an important point
                                                     needs to be made.
                                                       There are two dimensions to information: (1) facts and (2) under-
                                                     standing.  For  issues as  complex as those surrounding hazardous
                                                     waste, facts alone are seldom adequate to meet the public's need to
                                                     be informed. The  subject is fraught with complexities, limitations
                                                     and unknowns that require explanation.
                                                       The following example demonstrates the possible gulf between
                                                     facts  and  understanding. A groundwater monitoring program is
                                                     established, and residents are told their wells are being tested for
                                                     possible contamination. After a while, they are mailed results that
                                                     are, for all practical purposes, in an indecipherable code of initials
                                                     and numbers.  They are perhaps told that further testing may be
                                                     performed. For now, however, there is no immediate cause for con-
                                                     cern,  since no volatile organic chemicals exceed the state  action
                                                     level of 5 jig/I,  based upon a 10~' cancer risk level.
                                                       These are the basic facts, but these residents can hardly  be called
                                                     well informed. More likely, most of them are unable to derive any
                                                     true understanding from this information. As a result, they cannot
                                                     determine what the data mean to them or use the information to
                                                     make meaningful  decisions.  They may decide not to worry or to
                                                     drink bottled water as a precaution. Or they may decide to organize
                                                     their neighbors to demand immediate cleanup action.
                                                       The process  of risk assessment offers another illustration.* Esti-
                                                     mating human health risks from exposure to certain levels of con-
                                                     taminants contains a great deal of uncertainty.  The public  would
                                                     like to believe there are experts who can determine these risks with
                                                     a large degree of accuracy. Estimates of risk, however, are  based on
                                                     reasonable assumptions subject to change at any time. Numbers
                                                     and values are assigned, and regulatory decisions are made, but
                                                     (hey are not based on definitive scientific facts. In addition, risks
                                                     can never be reduced to zero. Choices  sometimes have to be made
                                                     —between jobs and health impacts, for instance, or between eco-
                                                     nomic development and increases in pollution.
                                                       The public needs to recognize the tradeoffs and limitations. But
                                                     the public should also understand the value of attempting to quan-
                                                     tify risks and the reasonableness of the process. Half explanations,
                                                     apparent contradictions and obscure processes will result in  a con-
                                                     fused, suspicious public. Only fully informed citizens will be able to
                                                     participate in complex decision-making.
368
PUBLIC PARTICIPATION

-------
CK^^aa AINU BALANCES

  Distinguishing between fact and understanding has an additional
benefit: it acts as a system of checks and balances for those respon-
sible for hazardous waste activities. Information that is meant to
convince the community that disposal and cleanup can be satis-
factorily accomplished must, in fact, be convincing. To describe a
process,  it must first  be examined; to explain a decision, it must
be reasoned. The act  of writing—or the thought of speaking at a
public meeting to people who are demanding to know—can serve as
an impetus toward clear thought and reexamination.
  The goal of genuine understanding also makes it easier to pro-
vide information  that is appropriate. The  basic  question  that
should be asked is: What does the public really need to know in
these circumstances? Answering this question requires listing to the
community  carefully, hearing not only the words, but also the
fears, frustrations and anger that may be behind them. It requires
measuring information already given to the  community against
community  perceptions of this information.  It means replacing
one-way assumptions  with two-way communication. In the end, it
helps prevent the costly error of offering information that  is off-
target, inadequate or insulting.
 ISSUES AND INFORMATION NEEDS

  With these thoughts in mind, some major issues and information
 needs are presented.

 Health Effects and Risk Assessment

  Some of  the most fundamental issues that arise concern the
 possible health effects associated with hazardous wastes. They are
 also some of the most complicated. As  discussed previously, so
 much is still unknown about the risks associated with many con-
 taminants. Yet, decisions  must be made, and their basis must be
 explained to the public. Health issues are further complicated by
 the emotion that often surrounds them. People's fears  are very
 real, even when they do not seem to be based on "fact."
  This was clearly demonstrated at a Superfund site in California.
 Residents near the site were extremely concerned about what they
 perceived to be an abnormally high  incidence of cancer and mis-
 carriages in their community. A number of people had also been
 complaining of headaches, nausea and vision problems for a num-
 ber of years. A local health clinic screened 75 residents, and the
 county  health department examined the findings. The  county re-
 ported that "the observed  morbidity of residents was not out of the
 ordinary and that no observed occurrences were statistically signif-
 icant." Residents simply  did  not believe  this conclusion. They
 organized to press for further health screening, which is now being
 conducted by the state.
  Discussions with community members  clearly showed why they
 would not accept the initial conclusion. They felt the information
 they had received was unclear and inconsistent. They did not re-
 ceive adequate or timely responses to their questions,  and the lang-
 uage of scientific detachment that was used  to give them the results
 was totally inappropriate to their needs. The current health survey
 will inevitably have the same credibility gap unless it profits from
 these lessons.
  It is  important that regulatory agencies and industries work
 closely with health professionals to develop materials about health
 effects and  risk assessment. The answers must be not only tech-
 nically sound and thorough, but must also be understandable to
 the lay public. Because these issues are so volatile,  the public may
 still not believe certain findings. Studies may still be  challenged if
 they show no connection between perceived health problems and a
 site that looks and smells terrible, but the chances of public accep-
 tance can certainly be increased by improved communication.
•The discussion about risk assessment is based upon: Rifkin, E., "Assessing Risks for the Biscayne
Aquifer System." Remedies: An Update of Hazardous Waste Issues. Biscayne Aquifer Project,
USEPA Region IV, Atlanta, GA., July 1984.
Costs and Benefits

  Another major concern of community members is the impact of
new facilities or uncontrolled hazardous waste sites on their prop-
erty. They fear that existing or potential contamination may lower
property values, make it difficult to sell or put restrictions on devel-
opment. While these costs are probably of greatest concern, other
costs may also be associated with hazardous waste sites: bottled
water must be used if wells become contaminated;  medical costs
are incurred if health effects are  experienced. At one Superfund
site in Colorado,  the fire district required several businesses to in-
stall methane detection systems because the businesses bordered a
landfill from which methane gas was migrating.
  People question why they must  pay  for  the  consequences of
someone else's actions.  In some cases, these costs can have a sig-
nificant impact on their lives. Again, the answers may not always
be wholly satisfactory. But  the questions warrant serious attention
and thorough explanation.
  A new hazardous waste disposal facility can  also offer eco-
nomic benefits such as jobs and training programs, tax revenues
and local purchases of supplies and  equipment. These positive im-
pacts should also  be understood by a community involved in siting
issues. They can then be weighed against the possible negative im-
pacts, enabling the making of reasonable choices.

Groundwater

  Groundwater is a complex technical subject. It is  also at the
heart of many hazardous waste incidents. In a USEPA study of
881 sites evaluated through  the hazardous ranking system, 60% of
the sites involved releases of contaminants into the groundwater,
with a potentially exposed population of 8.2 million people. Some
years ago, many people probably did not even know that ground-
water was their drinking water source; today, the increasing threat
to this source has resulted in a much greater awareness. How-
ever, this awareness does not always extend  to an understanding
thorough enough to address the issues.
  Residents may not understand how one well can be highly con-
taminated while adjacent wells are clean. Terms such as "plume"
and "aquifer" are still  foreign to many people. And the  public
probably does not yet realize its contribution to groundwater con-
tamination through small releases  of hazardous materials. Educa-
ting the  public about this subject  can help people understand
groundwater problems. It can also help prevent them.

Contaminant Standards
  Contaminant standards—or the lack thereof—are often not well
understood by the public. Terms such as "maximum contaminant
level," "primary  drinking water regulations" or "state action lev-
els" are confusing. People may not understand how allowable con-
taminant levels or cleanup criteria are determined, particularly in
cases where no enforceable standards have been established. These
issues are closely tied to risk assessment and suffer some of the
same complications. As with risk assessment, however, the ration-
ale and limitations need to be explained as completely as possible.

Safety and Reliability

  A number  of issues are incorporated within this general head-
ing. Common to them all is the public's desire to know what risks
and assurances are associated with siting or remedial actions.
•Technical processes:  How does a proposed technical process
 work?  What can go wrong; what are the chances of this happen-
 ing? Are there any by-products that could have negative impacts?
  The public is participating in more and more sophisticated eval-
 uations:  air stripping versus carbon adsorption; off-site disposal
 versus on-site containment; the reliability of proposed new facil-
 ities. Appropriate levels of technical background must be pro-
 vided to assist in  these evaluations.
•Regulatory Safeguards: What are  the  regulations  for site con-
 struction, operation  and closure?  What safety  measures,  emer-
 gency procedures and contingency plans will be provided?
                                                                                          PUBLIC PARTICIPATION
                                                          369

-------
  It may not always be adequate to offer assurances that local,
 state and  Federal standards are in place and must be met. The
 public may want to know exactly what the standards are, on what
 they are based and how they will be fulfilled.
•Oversight and Enforcement: Who is responsible for ensuring that
 regulations are being met and that enforcement will be  carried
 out?
  Past performance is a crucial element here. The credibility of the
 responsible authority must  be established  before public accep-
 tance will occur.

Programs and Policies
  Questions often arise about hazardous waste programs and pol-
icies. Funding priorities and schedules for Superfund remedial ac-
tions are generally not well understood. The definition and legal
liability of responsible parties also  need  to be clarified in many
cases.  Low-income communities sometimes perceive that sites in
their neighborhood are given lower priority than sites in  high-in-
come areas. Residents buying bottled water or suffering decreased
property values wonder why they are carrying the burden instead of
the waste  generator. These concerns must be addressed and can
sometimes be explained only within the broad context of regulatory
programs  and  policies. Good explanations may not always  be
forthcoming. If so,  this  is also  a legitimate topic of  discussion.
Limitations and omissions must be examined if the public is to be
truly involved in guiding public policy.

PUBLIC INFORMATION MATERIALS
   In developing public information materials, some basic guide-
lines should be kept in mind:
•Materials should contain the information  that is  needed. This
 means the audience must be clearly defined and its concerns must
 be identified.  While this advice may  seem obvious, past exper-
 ience shows it has not always been followed.
                                                        •Information should match the audience's level of concern. Ma-
                                                         terials should be neither condescending nor inappropriately elab-
                                                         orate.
                                                        •Information should be well organized and clearly presented.
                                                        •Information should be timely and accessible. The public should
                                                         know when and where it is available.
                                                        •Technical, legal and other specialized terms should be avoided
                                                         if they are not necessary and well explained if they are used.
                                                        •Slick, flashy materials are not appropriate. At best, they are un-
                                                         necessary; at worst, they can be perceived as a public relations
                                                         ploy and a waste of money.
                                                        •Some thought should be given to what vehicles best convey the
                                                         information. Fact sheets can be used to explain the various as-
                                                         pects of a project and  to provide periodic updates. Newsletters
                                                         may be appropriate for providing ongoing information to a large
                                                         audience. Where complex issues are involved and more thorough
                                                         public education is desired, issue papers, technical summaries or
                                                         question and answer sheets should be considered. In some cases,
                                                         public needs can best be addressed at public meetings that allow
                                                         two-way communication. Finally, graphic materials can be an
                                                         effective supplement to written or oral presentations. A diagram
                                                         that explains a technical process can greatly enhance a technical
                                                         summary. A slide presentation  at a community meeting can give
                                                         people  an accurate mental image of a proposed  project.  Like
                                                         written materials, graphic  materials should be clear, well pre-
                                                         sented and appropriate to the audience.
                                                        CONCLUSIONS
                                                          The public  will continue to  be involved in the siting, expansion
                                                        and cleanup of hazardous waste facilities. Past experience has iden-
                                                        tified a number of major issues that commonly arise and must be
                                                        addressed. Government and industry should provide information
                                                        to help the public understand the complexities surrounding haz-
                                                        ardous waste activities. Only a  truly informed citizenry will be able
                                                        to participate in responsible decision-making.
370
PUBLIC PARTICIPATION

-------
               COMMUNITY RELATIONS  ACTIVITIES  FOR
                ENFORCEMENT-LEAD SUPERFUND  SITES

                                             DAPHNE GEMMILL
                                   U.S.  Environmental Protection Agency
                               Superfund Community Relations Coordinator
                                               Washington, D.C.
                                        BRADLEY R. BROCKBANK
                                                ICF  Incorporated
                                               Washington, D.C.
INTRODUCTION

  A potential obstacle to success in the Superfund program can
arise when there are differences between the results of government
response and enforcement efforts and expectations of citizens con-
cerned about sites. Although a remedy may meet all the legal and
technical criteria established for the Superfund program, it may
still be perceived as inadequate by citizens living and working near
the site.
  The objective of the Superfund community relations program is
to minimize or avoid this obstacle  to success by creating a con-
structive exchange of information between citizens and government
response and enforcement staff. Through this exchange, citizens
can learn about the capabilities and limitations of the Superfund
program, and program staff can consider and incorporate, where
possible, the concerns of citizens into  their remedial and enforce-
ment efforts.
  In this paper, the authors explore the role of citizen participa-
tion during the Superfund enforcement process and describe sev-
eral mechanisms for including citizens in enforcement delibera-
tions. Except where noted, the views expressed are the authors'
and do not represent USEPA policies.
  The USEPA has established requirements and procedures for en-
suring that  the public receives information about site problems
and cleanup efforts  and has opportunities to comment on pro-
posed response actions.  The USEPA must develop a site-specific
community  relations plan for every Superfund-financed response
action. These plans must be based upon personal, on-site discus-
sions with concerned residents, citizen group  representatives and
local government officials in the site area.*
  The challenge to the  community relations program  is greater,
however, when the USEPA is conducting  enforcement actions to
secure cleanups by potentially responsible  parties.  Confidentiality
is often essential to successful enforcement efforts during or  be-
fore negotiations or litigation with  parties potentially responsible
for the release of a hazardous substance. Confidentiality is  im-
portant to ensure that civil cases are  not pre-tried by the public and
the media and to encourage candid and open exchanges of informa-
tion between potentially responsible parties and state and federal
government  enforcement  personnel.  Because the government's
negotiating position is greatly influenced by the strength of its case
against responsible parties, inappropriate disclosures of informa-
*For more details on the requirements and techniques of the Superfund Community Relations pro-
gram, see "Community Relations in Superfund: A Handbook, Interim Version", September,
1983, USEPA.
tion may undermine efforts to secure proper cleanup at a site. In
practice, these concerns often mean that the USEPA conducts en-
forcement-lead responses largely or wholly removed  from public
view.
  Residents in  the vicinity of a Superfund site, however, have a
legitimate interest in the outcome of a Superfund response action
regardless of whether that response is a Fund-financed govern-
ment cleanup or an enforcement action. It is the citizens—not the
government or  the potentially responsible parties—who must con-
tinue to live and work near the site long after the completion of
response actions. When citizens perceive their mental and physical
health to be at stake, they can be expected  to demand that their
interests be defended. Therefore,  when enforcement efforts are
conducted in confidentiality between potentially responsible parties
and the government, citizens are naturally inclined to be skeptical
of cleanup  decisions  because they perceive  themselves to be ex-
cluded.
  Can citizens be included in the Superfund enforcement process
in an active, meaningful way without jeopardizing the legal rights
of potentially responsible parties and the government's chances
for settlement? In most  cases, they can. Moreover, both govern-
ment and private parties involved in enforcement actions have an
obligation to encourage public participation in response decisions
affecting the community. Community relations activities may be
performed at six points during Superfund enforcement actions:
•Prior to the remedial investigation and feasibility study
•During and upon completion of the  remedial investigation and
 feasibility study
•During and upon completion of negotiations  with potentially
 responsible parties
•During and upon completion of litigation
•During responsible party cleanup
•During removal actions
  The USEPA's policies are not binding  on states unless the
USEPA is funding state-lead activities.  So far, the USEPA has not
funded any state-lead enforcement activities due to constraints
imposed by the Superfund law. However, the USEPA has recently
interpreted the  law to allow funding assistance for remedial inves-
tigations and  feasibility studies at state enforcement-lead  NPL
sites, and will begin to provide such assistance in fiscal year  1985.
USEPA policies regarding community relations during the RI/FS
will apply  in these instances.  If the law is amended to allow a
broader range  of enforcement funding support, then other pro-
visions of USEPA community relations policy can be made appli-
cable to state enforcement actions.
                                                                                    PUBLIC PARTICIPATION
                                                                                                                   371

-------
COMMUNITY RELATIONS PRIOR TO THE RI/FS

  Remedial enforcement actions can benefit greatly from an under-
standing of the concerns of the  local  community near a site.
USEPA policy requires the development of a community relations
plan based upon on-site discussions for all Super fund enforcement
actions. On-site discussions  provide an opportunity for USEPA
staff to:
•Learn  about the history of the site from the unique perspective
 of citizens living and working in the vicinity
•Gauge effectively the level and nature of citizen concerns about
 the site and their expectations for cleanup
•Determine the techniques for providing information to the public
 most preferred by citizens  and the issues about which citizens
 wish to have input
•Identify any other issues or  information about which the USEPA
 may be unaware
  These discussions may be particularly helpful to enforcement
efforts. Residents familiar with  the history of the site  may help
strengthen the government's case by providing information about
the past or present actions of responsible parties.
  A USEPA policy issued in January, 1984 should enhance the
value of on-site discussions  in yielding information about poten-
tially responsible parties. Under this policy, USEPA staff may re-
lease the  names of potentially  responsible  parties to the public,
either in response to a Freedom of Information Act request or on
the Agency's initiative. USEPA staff must inform citizens that the
parties' liabilities are  not proven, but that  those parties have  re-
ceived or may receive  notice  letters informing them of their poten-
tial liabilities. Releasing these names may prompt citizens to pro-
vide important information about the parties that they might other-
wise have considered unimportant to Agency enforcement efforts.
  There is another, perhaps  more important, reason for conduct-
ing on-site discussions with  citizens early in an  enforcement-lead
response. Government negotiators should have a clear understand-
ing of the expectations of local residents for cleanup at a site before
entering negotiations with potentially responsible parties. Those ex-
pectations should then be considered when the government devel-
ops its negotiating strategy. Although negotiated cleanups may not
always  meet citizens'  expectations, public understanding of such
settlements can be better achieved if government negotiators have
demonstrated a willingness to consider public concerns in preparing
for enforcement action.
  By conducting on-site discussions early in  the response, the
USEPA can develop a community relations plan tailored specifical-
ly to the information  needs and concerns of local citizens. For en-
forcement-lead sites, the plan should be developed in consultation
with legal and technical enforcement staff to ensure that the sched-
ule  of community  relations activities  is  consistent  with  the
schedule for enforcement actions. The plan should also contain
provisions for  a routine review process so that the appropriate
enforcement officials  can approve all information to be released
during the response.

COMMUNITY RELATIONS DURING AND UPON
COMPLETION OF THE RI/FS

  Once the remedial  investigation begins, community  relations
activities should proceed in accordance with the community rela-
tions plan. It is entirely appropriate, in most instances, to hold pub-
lic meetings, small group meetings, workshops,  to conduct other
informational activities to respond to public inquiries and to dis-
cuss site conditions, findings of studies and alternative remedial
actions under consideration. To avoid undermining possible future
enforcement efforts, however, the USEPA must refrain from dis-
cussing preferences for a particular remedy, the Agency's enforce-
ment strategy or the attitudes or positions of potentially respon-
sible parties.
  If the USEPA is in  active  litigation during the remedial investi-
gation and development of the  feasibility study, community rela-
tions activities are subject to the approval of the  Department  of
                                                       Justice and the courts. In addition, if USEPA officials believe lit-
                                                       igation will occur later in the response, legal enforcement staff
                                                       may decide to place some limitations on the information released
                                                       during the remedial investigation and feasibility study. The govern-
                                                       ment may discuss site conditions and the status of response efforts,
                                                       but interpretations of technical data  may need to be avoided in
                                                       order  not to put Agency or state officials on the  record in a way
                                                       that may jeopardize future litigation strategies.  Whenever such
                                                       confidentiality is necessary, however, the USEPA should explain
                                                       fully why it is necessary.
                                                         When the feasibility study is complete, it should be released for
                                                       public comment, unless  litigation concerns dictate otherwise, so
                                                       that citizens have an  opportunity to examine and  critique the
                                                       study's evaluation of alternative response measures.

                                                       COMMUNITY RELATIONS  DURING AND UPON
                                                       COMPLETION OF NEGOTIATIONS

                                                         Negotiations are one of the most sensitive aspects of an enforce-
                                                       ment-lead response and, therefore, the least  amenable to public
                                                       participation.  This should not mean, however, that contacts with
                                                       the community  must  cease  while the USEPA  negotiates  with
                                                       responsible parties. To most citizens, the negotiating process is a
                                                       mysterious "black box"  from which settlements emerge that may
                                                       or may not meet their expectations. It is perfectly reasonable to ex-
                                                       plain to citizens in general terms  how negotiations are conducted
                                                       and why it is important to ensure a measure of confidentiality. Cit-
                                                       izens should be  fully informed of the generic  issues that  are dis-
                                                       cussed during negotiations in addition to the technical aspects of
                                                       the remedy, such as releases from liability for work completed and
                                                       penalties for noncompliance.  Any release of information during
                                                       negotiations must  be approved by the appropriate enforcement
                                                       officials, however, in accordance with the review procedure set out
                                                       in the community relations plan.
                                                         The USEPA's experience shows that citizens respond favorably
                                                       when  the mystery surrounding negotiations is removed and the
                                                       USEPA staff  are straightforward about what  can and cannot be
                                                       revealed. At one Superfund site, for example, USEPA enforce-
                                                       ment and  community  relations staff  met regularly  with  citizens
                                                       throughout the course of negotiations. When citizens asked ques-
                                                       tions  which were too sensitive to answer, the  USEPA staff  hon-
                                                       estly stated their inability to answer. Ultimately, the citizens came
                                                       to trust the USEPA's responses and understood that certain issues
                                                       could  not be  discussed publicly. At the same time, enough ques-
                                                       tions  could be answered  to assure the citizens that their interests
                                                       were  safeguarded. In addition, the USEPA staff convinced the
                                                       potentially responsible party to meet with citizens to hear their con-
                                                       cerns directly. The company also attended other meetings to pre-
                                                       sent its study  proposal and to accept comments,  many of which
                                                       were incorporated into the final agreement.
                                                         A public comment period on the final agreement should always
                                                       be conducted. Administrative consent orders will contain a stipula-
                                                       tion that public comments may result in modifications  to the
                                                       order. When the comment period is over, the USEPA should pre-
                                                       pare a responsiveness summary to be sent to the appropriate Reg-
                                                       ional official,  who will then recommend either that the order be
                                                       signed without change or that negotiations be reopened to consider
                                                       the issues raised by citizens. In  any case, the order does not go into
                                                       effect until it is either signed unchanged or modified and approved
                                                       by USEPA negotiators and responsible parties. Those provisions
                                                       unaffected by potential changes may, however, be  implemented
                                                       without delay.

                                                       COMMUNITY RELATIONS  DURING AND UPON
                                                       COMPLETION OF LITIGATION

                                                         Litigation may or may not be initiated at a given site and,  if in-
                                                       itiated, may occur at any point  during an enforcement-lead re-
                                                       sponse. Community relations activities conducted after a complaint
                                                       has been filed in a federal district court must  be  approved in ad-
                                                       vance by the Department of Justice of the Assistant U.S. Attorney.
                                                       If a case is being tried in court, the judge will often  have final
 372
PUBLIC PARTICIPATION

-------
authority to decide whether community relations activities may be
conducted. Thus, it is inadvisable to provide contingency plans for
community relations  activities during litigation in the initial com-
munity relations plan developed prior to the remedial investigation
and the initiation of an enforcement action. A better approach is
to revise the community relations plan once the USEPA  refers a
case to the Department of Justice.
  If litigation results in a consent decree for site cleanup,  the De-
partment of Justice conducts a public comment period on the terms
of the decree, consistent with Department procedures, and pre-
pares  what amounts  to a responsiveness summary for the court.
The Department may seek to  amend the decree based upon the
comments received, or the decree  may be signed by the court and
take effect without change.

COMMUNITY RELATIONS DURING A RESPONSIBLE
PARTY CLEANUP

  Even when enforcement activities are successful and result in a
cleanup managed  by responsible parties,  the USEPA  remains
responsible in the eyes of  the public for the technical results
achieved. Therefore, the USEPA should ensure that adequate com-
munity relations activities are conducted throughout the design and
construction of the remedy. In some cases, the responsible party
may participate actively in community relations activities, but the
USEPA should provide careful oversight in such instances. The
appropriate roles of responsible parties in conducting community
relations activities should be determined on a  case-by-case  basis,
taking into consideration past experiences with the responsible
parties. For example, responsible parties who enjoy community
support may be capable of effectively conducting community re-
lations activities.

COMMUNITY RELATIONS DURING REMOVAL ACTIONS
   During removal actions, enforcement activities and community
relations activities are usually  significantly reduced in scope and
duration. The USEPA generally does not conduct negotiations or
initiate litigation for removals, unless there is sufficient time before
site action  must begin. The USEPA may, however, issue a uni-
lateral administrative order to compel responsible parties to take
discrete actions to abate an immediate threat, or arrive at an agree-
ment with the responsible party to undertake the work. This agree-
ment would be embodied in an administrative consent order.
  Once issued, a unilateral administrative order or a consent order
becomes a public document and should be made available for pub-
lic inspection. The USEPA community relations staff should be
prepared to discuss the  terms  of  the order and respond  to any
questions  or concerns raised by citizens. If the responsible party
fails to comply with the  order,  however, the appropriate enforce-
ment  officials should  approve  all  releases  of  information to the
public in the event civil action may occur.
  If an administrative consent order is issued for a longer term re-
moval action, the USEPA should  conduct an abbreviated,  two-
week  public comment period consistent with  USEPA policy on
initial remedial measures.

CONCLUSIONS
  In conclusion, the USEPA has an obligation to inform the public
and seek citizen input as much as possible during enforcement ac-
tivities at a site. Citizen awareness of hazardous waste problems is
at an all-time high, and citizens are continuing to demand more in-
volvement in decision-making.
  Community relations is particularly important for Superfund en-
forcement actions because  the  confidentiality that in many in-
stances necessarily surrounds such actions can  reinforce, rather
than diminish, citizen skepticism.  Retreating from  contacts  with
citizens is the worst  response because it erodes credibility  even
further. Instead, community relations during enforcement actions
should be approached with the knowledge that a successful en-
forcement action demands a clear  understanding of citizen  con-
cerns  and  expectations and that citizen input can effectively com-
plement enforcement efforts.
  In some instances, certain information may be judged too sensi-
tive to be released to the public. Nevertheless, only by listening to
citizen concerns and responding honestly to them—even if respond-
ing honestly means admitting that certain information is confiden-
tial—can the USEPA earn and maintain the public trust.
                                                                                         PUBLIC PARTICIPATION
                                                         373

-------
       MANAGING  CONFLICT IN CONTROVERSIAL SITING
                         ISSUES: THE KEYSTONE  PROCESS

                                          NANCY RYBURN WORST
                                                   EMPAK  INC.
                                                  Houston,  Texas
                                             DIANE B.  SHERIDAN
                                       League of Women Voters of Texas
                                           Taylor Lake  Village, Texas
                                                JOHN EHRMAN
                                               The Keystone Center
                                                Keystone, Colorado
INTRODUCTION
  Public opposition to new waste management facilities continues
to be the most important factor in unsuccessful siting attempts.
In recognition of the powerful influence of local residents and the
general public in permitting procedures for these facilities, many
states have devised procedures and formal structures to allow for
citizen input. The State of Texas has opted to use  the Keystone
Siting Process. In this paper, the authors describe public participa-
tion processes tried in other states and the Keystone Siting Process.

PUBLIC PARTICIPATION IN OTHER STATES
  The approaches taken by other  states to the public participa-
tion problem fall into four major categories:
•Private negotiation with a local committee
•State ownership of disposal site
•Approval of site by state siting board
•Approval of site by local committee or council
Massachusetts
  Massachusetts has chosen the first course of action. The Massa-
chusetts Hazardous  Facility Siting Act went into effect in 1980.  It
represented  one of the first efforts  to institutionalize the concepts
of negotiation and compensation into the hazardous waste siting
process. The Act calls  for  negotiation between the community
where a site is to be located and the applicant in an effort to de-
termine what compensation and/or mitigation might be necessary
to make the development acceptable to the community.
  Specifically, the Act created a 21  member siting council made up
of state officials, environmental and public interest  representa-
tives, scientists, other interest groups and  the general public. The
council oversees the functioning of the siting process which is in-
itiated by a  developer when a site  is proposed. The board  makes
a preliminary  decision  on the application to eliminate  frivolous
proposals and then oversees the creation of a local  assessment
committee,  a  local  group which will participate  in negotiations
with the developer. The local assessment committee has the power
and the duty to represent the host community in negotiations with
the developer, to negotiate specific mitigation measures and ob-
tain compensation for remaining adverse  effects,  to enter  into a
binding siting agreement and to appoint two local residents to serve
as voting members of the siting council.
  To date,  four applicants  have filed notices of intent with the
siting council. None of these applicants  has successfully moved
through the process and, in fact,  three of them have withdrawn
their applications. The fourth application  is still in process. None
of the four  applications has reached the negotiation stage in the
                                                    process. Those who have been observing the process have specu-
                                                    lated on why this is the case,  and it appears to be the result of a
                                                    number of factors, including:
                                                    •The process calls for a preliminary screening of a notice of intent
                                                     by the siting council to determine whether a proposal should enter
                                                     the process. Fifteen days were provided for the council to make
                                                     its determination. Communities, however, have pressured the
                                                     board to make substantive determinations on the merits of the
                                                     project in this period. This pressure has held up the progress of
                                                     the applications and led to solid positions being taken by the pub-
                                                     lic long before sufficient information is available.
                                                    •A second problem has been the fact that it is very difficult for the
                                                     state (or any portion  of the state bureaucracy)  to remain truly
                                                     neutral in the negotiation process. The siting board has to make
                                                     substantive and procedures decisions such as determining which
                                                     communities should receive technical  assistance  grants available
                                                     under the Act. Not surprisingly, the board has come under heavy
                                                     criticism for the decisions it has made.
                                                      Overall, the Massachusetts siting approach has not been as suc-
                                                    cessful as those who conceived it had wished. The realities of siting
                                                    facilities present a great challenge even to those processes designed
                                                    to be responsive to the needs of the various parties concerned with
                                                    siting decisions.

                                                    Wisconsin
                                                       Wisconsin's  plan is  another variation of the  private  negotia-
                                                    tion with a local committee option. In May, 1982,  a major revision
                                                    to state law in Wisconsin established a mechanism for negotiations
                                                    between  the prospective operator of a  site and a  local  committee
                                                    representing affected municipalities. It also created a Waste Facility
                                                    Siting Board to arbitrate in case of impasse.
                                                       The negotiation/arbitration process was designed by an ad hoc
                                                    legislative committee representing the  state senate and assembly
                                                    and industrial, environmental, state agency, regional planning and
                                                    local government interests. The committee wanted to involve local
                                                    parties at an early stage in site development, but with a minimum
                                                    of delay  to the siting process. So,  under the law,  negotiations are
                                                    set in motion ahead of regulatory procedures conducted by  the
                                                    Department of Natural Resources and may continue, if necessary,
                                                    as an independent but parallel process.
                                                       Governmental entities that are eligible to appoint members to the
                                                    local committee are:
                                                    •Any town, village or city in which all or part of the facility will be
                                                     located (four members)
                                                    •Any county in which all or part of the facility will be located (two
                                                     members)
374
PUBLIC PARTICIPATION

-------
•Any town, village, city or county whose boundary is within 1,200
 ft of that part of the site where waste will actually be disposed,
 stored or treated (one member)
  The triggering mechanism for negotiations is passage of a siting
resolution by any  affected municipality. The siting resolution is
sent  to the Waste Facility Siting Board which then notifies the
other affected municipalities that they must pass a siting resolution
within two weeks  to  be eligible to participate in the local com-
mittee.
  Once all the siting resolutions have been passed and local com-
mittee members have been appointed, negotiations can begin. To
become  binding, all items agreed to by the local committee and
the applicant must be approved by the governing  bodies of the
communities in which all or part  of the facility will be  located.
Eligible  items on which agreement cannot be reached may be in-
cluded in an arbitration proposal to be decided by the Waste Facil-
ity Siting Board.
  Any item may be the subject of bargaining except the need for
the site  or any proposal  that would weaken environmental pro-
tection measures. The topics that may be arbitrated  are fewer and
are listed in the statute. If it believes further progress toward settle-
ment is possible, the siting board is empowered to order continued
siting negotiations.
  The Waste Facility Siting Board is a seven-member body com-
posed of representatives from four state agencies (Transportation;
Development;  Industry, Labor and Human Relations; and Agri-
culture, Trade and Consumer Protection) and three local elected
officials (two town and one county) appointed by the  Governor.
  Once the board accepts a  case for arbitration, it must choose
between two final offers presented by the parties. The final offers
may include only items that have been offered in negotiation and
are defined as arbitrable under the  law. The board must select one
of the final offers in total and may not select items from each. If the
board fails to decide by at least a five to two majority, the arbitra-
tion decision will be made by the Governor.
  The board had its first meeting in late August, 1982. It has not
yet received any petitions for arbitration. By mid-March, 1983, it
had 16 cases on file in which siting resolutions have been passed
and negotiations are anticipated or underway. One  petition for a
determination of default was filed, but it was  settled informally
before it reached the board. To date, no hazardous waste  facilities
have been proposed.

 Minnesota
   Minnesota has tried to pick the optimum location for a state-
 owned site. The facility is to be operated by a private company.
 The supposed advantage  of a state-owned facility is that the long-
 term responsibility for monitoring a site will rest with a govern-
 mental entity (the state) which is likely to exist and have the finan-
 cial resources to maintain the site many years after closure.
   The Minnesota  Board  members represent state and local inter-
 ests. Their selection  process for sites has narrowed the possible
 locations to four sites. Once the list of possible sites was narrowed
 to the final four, local opposition organized in each area. An effec-
 tive lobbying  effort by these groups resulted in  state legislation
 which has put a moratorium on choosing a final site for the last
 year and a half.

 Michigan
   In the State of Michigan, a Site Review Board is formed to con-
sider each site. The Board is made up of five permanent members
and four temporary  members. Of the five permanent members,
three are representatives of  state agencies:  the  Department of
 Natural  Resources, Department of Health and the State Police.
The other two permanent members are a geologist and a chemical
engineer appointed by the Governor. Of the four temporary mem-
bers of the Board, two are appointed by the governing body of the
municipality where the facility is to be located. The  remaining two
members are local residents appointed by the County Board.
  The Department of Natural Resources (DNR) chairs the Board.
One of the criticisms of this process is that the DNR is not seen as
being neutral in this process. The process has met with some success
however. Five facilities have been considered so far. The two facil-
ities which were  on the site of a manufacturing process were ap-
proved. The three  proposed  facilities which were to be developed
exclusively for  the  commercial  disposal  of wastes  were  not
approved.
Colorado

  In Colorado,  local counties must  approve or disapprove pro-
posed sites before they  are  permitted.  The political  pressure on
county  boards  has  prevented  any  proposed sites  from  being
approved.

KEY ELEMENTS  OF SITING APPROACHES
  A review of the  siting processes which have been developed in
Massachusetts, Wisconsin and elsewhere indicates that there are a
number of important factors  to consider in developing  a hazardous
waste facility siting process:
•There must be early and substantive involvement of the public
•The process must provide for a range of interests to have input
 into the siting process
•The parties involved in a  siting process  need scientific, tech-
 nical and procedural support
•The siting process must be efficient and implementable

THE KEYSTONE  SITING PROCESS
  The Keystone Siting Process is the product of participants in two
workshops conducted by the Keystone  Center in August and
October of 1982. The more than 30 attendees came from a diver-
sity of  backgrounds including industry, government,  environ-
mental groups, labor and civic organizations. The goal of the work-
shops was to develop a workable siting process for new hazardous
waste facilities in the Galveston Bay Area. The process should be
applicable to all  of Texas as well as to other states and other types
of siting efforts.
Advantages
  Establishment  of  a review committee to facilitate   dialogue
between applicant  and public is the heart of the Keystone  siting
process. Providing this less formal, prehearing forum  to raise and
perhaps resolve  issues of mutual  concern is advantageous  to all
those interested in a proposed facility: applicant, community and
permitting agency. The  review committee's goal is to develop a
report dealing with local citizen concerns and the manner by which
the applicant is  dealing  with those concerns. Advantages  of the
committee process are:
•Early citizen input is possible
•Nontechnical issues can be addressed
•Areas of conflict are identified and possibly resolved
•Reliable information is provided to the community
•An informal exchange of  information  takes place  between an
 applicant and the community
  The  review  committee provides a  means of dialogue between
the applicant and a duly constituted group representing commun-
ity  interests, thus eliminating the difficulty faced by an applicant
who wants public input  but  is not sure whom to contact. The re-
view process undertaken by the committee permits accommodation
of viewpoints  during a time period prior to review of the permit
application by the  respective state agency. This timing allows pub-
lic involvement in  the consideration of a proposed facility earlier
than is usually the case.
  By working with the public early in the siting process, the appli-
cant has a better chance of eliminating the  misinformation that is
often generated  about a project. The committee also offers both
the applicant and the public the opportunity to discuss social and
economic issues  that may not be admissible in a hearing but are
nevertheless of great concern.
                                                                                          PUBLIC PARTICIPATION
                                                         375

-------
  The review committee is able to express the concerns of the pub-
lic to the applicant, thereby providing the applicant with the oppor-
tunity to respond to  those concerns  to the extent possible. Re-
sponses might include furnishing data, making changes in the pro-
posed project and  making other accommodations which  might
alleviate some of the public's concerns.

Committee Review

  The committee is limited to reviewing a particular facility pro-
posed at a specific site. Committee time  and energies should not
be expended exploring alternatives to the  proposed  project, al-
though  members will want to know if the applicant explored such
alternatives.  The committee's charge  is not like  that  of siting
boards  which actually seek or choose sites for  hazardous waste
management facilities in some states.
   As an entity, the review committee is not to be a party to the pub-
lic hearing held by the permitting agency. Individual committee
members, of course, are not  precluded from participation  in the
hearing process, representing  viewpoints of their own or of organ-
izations with which they are associated.
   To initiate the Keystone Siting Process, the applicant indicates to
the regulatory agency that it  intends to file a permit application.
The applicant and the regulatory agency then contact those entities
involved in appointing committee members. The committee should
be selected within IS days and meet within 21 days  of the filing of
the notice of intent. The review committee  then has until the 90th
day to study the  proposal and write a report. The  process can be
extended by mutual agreement between the applicant and the com-
mittee.
   The applicant's representatives should  have sufficient responsi-
bility and authority to deal effectively with the issues raised and
make commitments on the part of the company. Failure to assign a
person of sufficient stature may  be perceived by the committee as
an indication of lack of real interest in the process on the part of
the applicant.
   To streamline  the information gathering and communications
process, the committee may expect,  in many cases,  general in-
formation on the facility proposal from the applicant in the earlier
stages of the process. This is to be followed by more specific tech-
nical  data as environmental and  engineering studies  proceed.
Throughout the process, all information available to the regulatory
agencies will be available to the committee. Questions and answers
during committee meetings can be staged from general to specific
to maintain a cooperative dialogue.
   After the filing of  the report,  the committee is  on "standby"
status to respond to changes proposed by the applicant and amend
its report as appropriate. If no amendment is made, the committee
is dissolved at the initiation of the public hearing. In no case is the
committee to participate in the hearing as a "party" to the pro-
ceedings, although individual committee members are  free to par-
ticipate in their own right.

Committee Support

   Since committee members serve as volunteers, expected expenses
are for clerical service and logistical support. However, an appli-
cant may wish to employ professional staff or a professional facil-
itator. It is also possible that  the applicant will want to reimburse
out-of-pocket expenses for committee member's mileage,  for ex-
ample,  though it is not anticipated that committee work will en-
tail significant expense on the part of the members. The commit-
tee will have no financial resources of its own, however, so cler-
ical support at a  minimum will  be needed. Large expenditures on
the part of the applicant would give some  the impression that com-
mittee members may become obligated to the applicant, so a bal-
ance must be struck whatever options are chosen.
   The committee develops Its report as it chooses, but the appli-
cant will at least  pay  for clerical help and printing. Copies  of the
report should be  made available  to the committee and the permit-
ting agency as well as local libraries and city halls for public use and
perusal.
                                                       Time Commitment
                                                         It is difficult to anticipate a committee's workload in terms of
                                                       hours. However, within the 90 days that the committee functions,
                                                       members can expect weekly-to-monthly meetings of two or three
                                                       hours in length. Members share the writing of the report unless a
                                                       staff person is  used to write the draft. In any case, members re-
                                                       view, comment on and edit the final report. Committee members
                                                       can expect to spend some time on this process; they  are not meant
                                                       to be mere window dressing.
                                                         The review committee is constituted to be: (1) open to fact-find-
                                                       ing, not locked into positions by external forces such as having
                                                       financial interest in the project or being an elected official in a town
                                                       near the site; (2) representative of a breadth of interests; (3) reflec-
                                                       tive of local concerns; and (4) responsive to its charge of prepar-
                                                       ing a report. To meet these requirements, it is appointed by, but
                                                       does not include, elected  officials;  it includes, but is not domi-
                                                       nated by, persons directly affected by the site; it reflects a range of
                                                       interests and expertise, such  as engineering, business, environ-
                                                       ment;  and it is limited  to individuals in the general region  sur-
                                                       rounding the site. In order that the committee be perceived as  rep-
                                                       resenting the public, most members need to be known to the "con-
                                                       stituency" they represent. For example, the person representing en-
                                                       vironmental interests should be known to the  environmental com-
                                                       munity.
                                                       Committee Membership
                                                         The committee should  be as  small as reasonable, given these
                                                       principles, to facilitate discussion and the preparation of a report.
                                                       Given the size, the need for representation and the need to have en-
                                                       tities do the appointing, the committee is designed as follows:
                                                         Eight regional members appointed by  a  regional entity  and
                                                       representing various interests:
                                                       •Environmental Groups
                                                       •Academia
                                                       •Industry
                                                       •Community Planner
                                                       •Public Interest
                                                       •Medical
                                                         Four members, living within five miles of site, appointed by local
                                                       mayor(s) and county judge.
                                                         A 12 member committee is best suited to developing a written re-
                                                       port. Four members come from the exact locality of the project and
                                                       eight from a somewhat wider region. The ratio between committee
                                                       members  from the immediate vicinity of the proposed  site  and
                                                       those from the region is critical  in maintaining a balance of con-
                                                       cerns and opinions and, therefore,  the credibility of the commit-
                                                       tee. A  committee predominantly local would not be seen by the
                                                       applicant  and  industry in general as being objective. A committee
                                                       without local residents would not be seen by the community as re-
                                                       flective of its concerns.
                                                       Report
                                                         One of the major tasks of the committee, and the  ultimate focus
                                                       of  its work, is preparation of a report detailing its work and sum-
                                                       marizing its findings. This report, which is submitted to the  per-
                                                       mitting agency simultaneously with the permit application, is not to
                                                       be  a recommendation of approval or disapproval of the proposed
                                                       facility. Other than that qualification, the scope of the report  is at
                                                       the discretion  of the committee as long as certain elements are in-
                                                       cluded. The report should  document discussion of community con-
                                                       cerns raised during committee review, including identification and
                                                       discussion of the following:
                                                       •Those issues which were resolved,
                                                       •Those issues which were not resolved,
                                                       •Those questions which were not answered, including why  they
                                                         were left unanswered.
                                                         In addition,  the report  explains why the committee was estab-
                                                       lished. Members are listed with their associations/backgrounds and
                                                       how they  were  appointed.  The procedures the committee followed
                                                       in performing its work are also described.
 376
PUBLIC PARTICIPATION

-------
  The committee report is a means of documenting the commit-
tee's fact-finding work and of providing information about  the
proposed facility and related  issues. It is not intended to be an
additional regulatory burden for  the applicant,  the permitting
agency or the community. Rather, an applicant who chooses to use
the review committee process will be creating an advantage to all.
  From the public's viewpoint, the report allows the permitting
agency to hear its concerns at the same time the staff is examining
the permit application. From the applicant's viewpoint, the report
shows that efforts were made to listen and respond to public con-
cerns early enough to make improvements to the planned facility.
From the permitting agency's viewpoint, the report provides a bet-
ter understanding of all  facets  of the proposed  facility and its
effect on the affected community.
  The use of the committee and its report should serve to improve
the present siting process. It incorporates the  four key elements of
siting approaches described earlier.  The public is involved early in
the process, even before  the regulatory agency review. A broad
range of interests is represented on the review committee and an
even broader range of input is derived from the general public
attending committee meetings.  The committee receives technical
and procedural support during its review of the project. Last of all,
this process is efficient (adding only 90 days to the permit process)
and implementable. No new laws or regulations are required to use
this process in Texas. Ultimately, the Keystone Process should help
ensure that the sites permitted in Texas are environmentally sound
and able to meet Texas' needs.

REFERENCES

Siting Waste Management Facilities in the Galveston Bay Area: A New
Approach. Report of the Keystone Workshop on Siting  Nonradioactive
Hazardous Waste Management Facilities. The Keystone Center, Keystone,
CO, 1983.
The Keystone Siting Process Handbook: A New Approach to Siting Haz-
ardous Waste Management Facilities. Texas  Department of Water Re-
sources and Texas Department of Health, Jan., 1984.
                                                                                           PUBLIC PARTICIPATION
                                                          377

-------
                  DO  COMMUNITY  RELATIONS  MATTER?:
                          THE NEW  JERSEY PERSPECTIVE

                                               GRACE L.  SINGER
                            New Jersey Department of Environmental  Protection
                                         Division of Waste Management
                                  Hazardous Site Mitigation  Administration
                                               Trenton,  New Jersey
INTRODUCTION

  Why should the Superfund Community Relations Program in
New Jersey be of interest to those outside the State?  There are
several reasons, but  the principal one is New Jersey's  aggressive
policy of identifying and investigating hazardous waste sites and
seeking Federal and State funding for remedial action. This policy
has put New Jersey at the top of the National Priorities List (NPL)
with 85 Superfund sites as of August, 1984 (in a national total of
546 sites) and a likely total of about  100 Superfund sites by the end
of this year. These figures do not include State funded sites which
should total about 75 in  1984. Michigan is second to New Jersey
with 48 sites on the NPL.
  The New Jersey figures, when added to those of other Northeast
states, two of which are third and fourth on the NPL, i.e., Penn-
sylvania with 39, and New York with 29 Superfund sites, emphasize
the regional context of the uncontrolled hazardous waste site prob-
lem in the nation's oldest industrial area. In addition, New Jersey
is the most densely populated and industrialized state in the nation,
and one in which the petrochemical industry is by far the largest in-
dustry.
  The State's coastal geography,  with porous, sandy soils  and
highly vulnerable groundwater which provides a substantial portion
of the population's potable supplies,1 is  another major factor
creating intense  environmental pressures and  projecting New
Jersey's hazardous waste issue to a prominent position. In an in-
teresting juxtaposition  to the industrial impression of  the State,
New Jersey's  two  second  largest industries  are tourism  and
agriculture, both requiring a clean environment in order to thrive.
With a high population density and many  sites in proximity to
population centers, citizens  of the state have  made cleanup of
hazardous wastes one of their greatest concerns.

CHALLENGES IN THE STATE PROGRAM

  As a result of all of these factors, New Jersey faces one of the
most complex and extensive hazardous waste cleanup challenges of
any state. To effectively manage the cleanup of the State's sites, the
New Jersey Department  of  Environmental  Protection (NJDEP)
developed a comprehensive management plan. The document, call-
ed the "Management Plan for the Cleanup of Hazardous Waste
Sites in New Jersey 1983-1986," outlines a systematic approach to
remedial action.2 New Jersey's Plan was the first such document in
the nation.  USEPA  is  now  encouraging other states  to develop
similar plans.
  Although public  hearings on the Plan  were not mandatory,
NJDEP held three such hearings last fall. After assessing extensive
public comment, the Department issued a Response Document to
                                                    clarify public policy and technical issues.' In addition to the ap-
                                                    proximately  175 sites to be addressed in the State's Management
                                                    Plan, now in revision. New Jersey has tentatively identified more
                                                    than 1,100 abandoned or improperly managed sites that are poten-
                                                    tially hazardous. This large number is primarily because the State
                                                    has one of the largest concentrations  of petrochemical and phar-
                                                    maceutical industries in the nation. In addition to the  threat of
                                                    human  exposure,  there  is great  concern  that improper  waste
                                                    disposal may be imperiling the State's ecological resources in-
                                                    cluding sensitive groundwater aquifers, the  unique Pinelands and
                                                    wetland areas, natural resources which must be protected. In fact,
                                                    New Jersey  has recognized the importance of potential natural
                                                    resources damage and has filed claims totalling $1.27 billion with
                                                    the Federal government under a special provision of the Superfund
                                                    Law.
                                                      Even before the  passage of Superfund,  New Jersey  began to
                                                    undertake hazardous site cleanups. Initial remedial actions at sites
                                                    such as Chemical Control, Goose  Farm and A-Z  Chemical were
                                                    completed by 1982 with state Spill Fund monies.4 Chemical Control
                                                    was  the largest hazardous waste  drum removal  operation ever
                                                    undertaken in this country.
                                                      With the passage of the Superfund  legislation. New Jersey was
                                                    able  to take advantage of prior information needed to qualify sites
                                                    for federal monies.  As a result, New Jersey has developed one of
                                                    the most advanced cleanup programs in the country.

                                                    ACTIONS TO DATE

                                                      To date, New Jersey has signed cooperative agreements or con-
                                                    tracts with the USEPA for approximately 30 sites to  complete
                                                    feasibility studies, design and construction activities. By the end of
                                                    Federal FY '84 or the beginning of FY  '85, 39 feasibility studies, 16
                                                    designs, 6 construction projects and 11 immediate removals expect
                                                    funding. In addition, the State has investigated 17 major potential
                                                    dioxin sites and has undertaken emergency dioxin cleanup/contain-
                                                    ment at four such sites.  PCB contamination, which posed an im-
                                                    mediate health threat in the City of Perth Amboy, was successfully
                                                    cleaned up in the summer of 1983. A  major cleanup effort under
                                                    Superfund at the Syncon Resins facility in Kearny was completed in
                                                    the third quarter of 1984. This project involved the removal and
                                                    disposal of almost 13,000 drums of chemicals.
                                                      By the spring of 1984, a total of 564 homes in northern New
                                                    Jersey were sampled for the presence of radon gas; 45 homes were
                                                    contaminated. Remedial actions have thus far been conducted at 22
                                                    of the homes.'
                                                      In 1983, the NJDEP completed the cleanup of 33 small drum
                                                    dump sites using funds from the State's Spill Fund; an additional
378
PUBLIC PARTICIPATION

-------
36 small drum dump cleanups should be completed by the end of
1984. Because of the variety and number of New Jersey sites, when
the State's experience is quantified it may establish a useful data
base for Community Relations as well as general remedial site ac-
tivities.

COMMUNITY  RELATIONS WITHIN THE
ORGANIZATIONAL STRUCTURE

  The Community Relations Program in Superfund cleanups  is
resource intensive  throughout  the  complex and often lengthy
cleanup process  (Tables  1 and 2). The Program presents a signifi-
cant departure from public participation as it has  been commonly
practiced under the National Environmental Policy Act (NEPA) of
1970  (amended  in  1975) and subsequent  Federal  and State laws.
This difference occurs because the nature of hazardous sites, often
literally in people's backyards or affecting drinking water, evokes
fears  and concerns not exhibited in response to more remote pollu-
tion threats.
  These concerns must be addressed as directly as possible, and the
Community  Relations Program is a  good vehicle  for  two-way
meaningful  communications between  citizens  and  government.
Even with the best of intentions, however, it is not usually possible
to accomplish site  remediation  with the speed and definitiveness
demanded by an upset citizenry and its elected officials.
                            Table 1
         New Jersey Department of Environmental Protection
                  Division of Waste Management
             Hazardous  Site Mitigation Administration

           A Community Relations Program at Superfund
                      Hazardous Waste Sites
   As part of the federal/state program of cleanup at hazardous waste sites,
 a Community Relations Program is conducted to receive local input and to
 advise local residents and officials about the planned remedial actions at
 the three major stages of the cleanup: (1) remedial investigation/feasibility
 study, (2) engineering design and  (3)  removal/treatment/construction.
 Local briefings and public meetings are conducted with elected officials and
 residents and  generally take place at:
 •The commencement of a remedial investigation/feasibility study to ad-
  dress local concerns early in the process
 •The completion  of a feasibility study to discuss the alternative courses of
  remedial action; there is a 30-day comment  period after public pre-
  sentation of the alternatives during which the feasibility study is avail-
  able in local repositories.
 •The engineering design stage to carry out the mandates  of the selected
  remedial alternative.
 •The completion  of the remedial action
   In  addition to the more formal activities outlined above, there is general-
 ly informal communication with local officials and residents. Depending
 upon whether the New Jersey Department of Environmental Protection
 (DEP) or the  USEPA has the lead in remedial action at a site, community
 relations activity  is conducted by the relevant State or Federal agency.
   In  New Jersey, the DEP Community Relations Program is conducted by
 Grace Singer, Community Relations Program Manager (609)  984-3141/
 4892. At Region II, EPA, the contact person is Lillian Johnson, Communi-
 ty Relations Coordinator (212) 264-2515.
   This is often due to the complexity of hazardous waste problems
 (which can vary widely from site to site), the number of sites to be
 addressed (especially in New Jersey), and a relatively new program
 which is often operating at the cutting edge of technology.
   Difficult problems often demand difficult decisions. In this con-
 troversial setting, it is critical to set out the goals of the Community
 Relations  Program  and understand the importance of public in-
 volvement. Clearly, in passage of laws mandating the Federal and
 State cleanup program, the public, through its elected officials, has
 already involved itself in a strong mandate for action.  Some of the
 key goals of Community Relations in Superfund can be clarified by
 asking the following questions:
 •How  will citizens' concerns  about  a site be identified, assessed
  and addressed?
                            Table 2
       Steps Involved in a Major Hazardous Waste Site Cleanup


 1.  Site Identified and Referred
 2.  Initial Site Investigation
 3.  Secure Site
 4.  Site Analysis Evaluation and Assessment
 5.  Prioritization
 6.  Determination of Lead
 7.  Community Relations Plan Activated
 8.  Signing of Contract or Cooperative Agreement
 9.  Hiring of Contractor for Remedial Investigation/Feasibility Study
10.  Preparation of Feasibility Study
11.  Selection of Remedial Action Alternative
12.  Hiring of Contractor for Engineering Design
13.  Hiring of Construction/Removal Cleanup Contractor
14.  Cleanup Evaluation
15.  Contractor Audit and Close Out
•How will accurate information on the problems associated with a
 particular site be explained and disseminated to local residents?
•How will the remedial alternatives and the proposed solutions
 be explained to the community?
•How will citizens have adequate opportunity to comment and
 provide input on ongoing site work before major decisions are
 made?
  In New Jersey, the State used these questions as guidelines and
criteria for Community Relations activities with generally favorable
results in a program still in its infancy.
  In order to meet the challenge of two-way communication with
citizens in the intensive New Jersey cleanup program, an Office of
Community Relations was established in January, 1983, within the
newly formed Hazardous  Site Mitigation Administration in  the
NJDEP (Tables 3 and 4). The Office of Community Relations has
five staff members:  a Community Relations Program Manager,
three  Senior Area Coordinators  assigned to specific sites and a
secretary.
  In a team effort,  the Community Relations unit works closely
with other units within the Hazardous Site Mitigation Administra-
tion (HSMA), especially the Bureau of Site Management and other
NJDEP  technical and  legal staff members. This  structure  has
fostered the team approach to public communications on site ac-
tivity. Technical staff as well as outside contractors conducting the
actual field work participate fully in  local  briefings and public
meetings with residents and officials. Thus, citizen questions and
comments can be answered directly by those conducting or manag-
ing site activity. As a result, such comments are heard with the full
flavor and fervor of deeply concerned citizens.  This method of
operation has conveyed a sense of urgency and added the human
dimension  which  can be  lost when dealing with  matters in a
technical framework only. Likewise, local citizens receive responses
directly from those conducting site work.
  Effective Community Relations in cleanup activities is seen  as
important  to the overall cleanup effort in New Jersey, and a con-
siderable commitment has been made to it. This is especially true
considering the number of New Jersey NPL sites; public meetings
take place at night when local residents and officials are available.
  Likewise, citizen comments are taken seriously.  One notable ex-
ample of this NJDEP response to citizen concern occurred recently
at a site in central New Jersey. Local residents pressed for the in-
stallation of a public water system  to replace their on-site wells
drawing from ground water contaminated by dumping at the site.
USEPA headquarters opposed the new water system, preferring to
continue groundwater and  well monitoring. The NJDEP agreed
with the residents,  whose comments and input to  the decision-
making process were rewarded with a reversal in the USEPA's deci-
sion.
  In another case, a local committee wanted to be closely included
in the initial stages  of the feasibility study because  members  felt
they had developed  special knowledge about the site. They asked
that the Request for Proposal (RFP) include a provision that the
                                                                                            PUBLIC PARTICIPATION
                                                           379

-------
 TJ
 00
 r
 n
 •o
o
13
>
o
z
                                                                                  COMMISSIONER
                                                                                  Robert E. Hughey
                                                         DEPUTY COMMISSIONER
                                                              Aiu/ Arbaman
                    OFFICBOF
              SCIENCE AND RESEARCH
                     r.
                         ASSISTANT COMMISSIONER
                         FOR NATURAL RESORUCES
                              Helen C Faatt
                                                                                         ASSISTANT TO COM.
                                                                                              A On*
                                                                      EXECUTIVE ASSISTANT
                                                                                            PLANNING
                                                                                             GROUT
                                                                                            L Schmidt
                                                                                             ASaStAMT
                                                                                              FUIRBGUAKKYAND
                                                                                             QUVHMBirALAfFAM
                                                                                                                                              r AfUtn
                                                          ASSISTAKT COMMISSIONER
                                                           KANACEMENT * MJDGET
       PARKS*
      FORESTRY
     FomiFinMfi.
      FomiMd.
      ini Ua* %1.
      yoMkStxtea
   IWU UW ErfQMMMM
GREEN ACEESA
 RECREATION
  M Ltdtma
                    SUM LM4 AcvMlkM
                     Laai GIMU
HSHCAMEA
 WILDLIFE
                 FmkwiM Fl
                 U*
                                       tmt. A
     TAL
RESOURCES
/<.
                                                      EWara. A FbU Sm.
  FBCALA
SUtOKTSERV
  r / MUw
                                                                                   ASSISTANT COMUBSIONEI FOR
                                                                                   ENV. MANAGEMENT A CONTVOL
                                                                                   	Gtoryl. Tfkr
                      WATER
                                                                                           •km QwMr M«i
ENVIRONMENTAL
   QUALTTY
   WASTE
MANACEMENT
                                                                                Table 3
                                                                   Department of Environmental Protection

-------
                                                             Table 4
                                              Hazardous Site Mitigation Administration
                                                       Administrator
                                                         Deputy Administrator
                 Administration
Bureau of Site Management
  A. Farro-Chief
  L. Romino
  G.King
                                                                                       Contracts Management
                                                                                         Joe Wiley—Chief


                                                                                       Community Relations
                                                                                         Grace Singer—Chief
Bureau of Environmental Evaluation
& Risk Assessment
  M. Morris—Chief
  S. Santora
  R. Predale
  J. Goliszewski
1
 Bureau of Site Operations
   R. Senna—Chief
   T. Allen
   F. Metzger
Bureau of Industrial Site Evaluation
  A. McMahon—Chief
R. Katz
 contractor must consult with them and review their extensive files
 before commencing the feasibility study. Because of their special
 experience with the site over a long period of time, this request was
 granted and the contractor did indeed work with the citizen group
 before proceeding with on-site work.
   Citizens can also thwart the  legitimate exchange of information
 between a government agency and local residents when it serves
 their purposes. One such incident occurred when the Department
 invited local residents  to a three-day public information "open
 house" at a dioxin site. It was apparent from the unusually low
 turnout of residents and a demonstration conducted by an organiz-
 ed local group that a show of dissatisfaction, rather than gaining
 firsthand information,  was a local priority.

 ISSUES AND QUESTIONS TO CONSIDER
   One overriding issue for consideration is adequate staffing and
 the extent to which the Community Relations function is genuinely
 integrated into the hazardous  site cleanup organization. Is Com-
 munity Relations considered a frill or is it taken seriously? Is the
 Community Relations  staff part of the internal decision-making
 process or is it seen as not being involved in the substance of issues
 causing public sensitivity? Part of the answer to this will be revealed
 in the background and abilities of the people selected to conduct
 the Community Relations program.
   Related to the above question is the issue of the extent to which
 there is  a team approach to Community Relations with technical
 and legal staff joining the effort for direct two-way communica-
 tions between government and the local community.
   To what extent will  Responsible Party or Private Party enforce-
 ment cases be involved in Community Relations? Based  on the
 premise that a local community has the right to know what is being
 done at a site which is causing adverse effects on public  health and
 the environment, regardless of who is conducting the cleanup, it is
 essential  to develop a workable process assuring effective com-
 munication and responsiveness to local citizen concerns in the
 Private Party  cleanup. Such  communication should be equal to
 that of government agency conducted cleanups.
   The always sensitive issue of land use is intensified when hazar-
 dous waste sites are involved.  Because local communities in most
 states have land use authority while State or Federal agencies have
 hazardous waste site data, there is a growing information gap. This
 is especially serious in developing communities where  residential
 subdivisions may be planned near hazardous waste sites. There ap-
 pears to be very little occurring in the transfer of information be-
 tween the  government cleanup agency  and the local land use
 decision-making body.
   Because of the magnitude of this issue in New Jersey, a Land Use
 Inquiry Program within the Hazardous Site Mitigation Administra-
                                    tion has been established to disseminate information to prospective
                                    home buyers and others. An average of 85 such calls are responded
                                    to monthly. In order to minimize future problems, a community
                                    may take precautions such as establishing a building  moratorium
                                    within a certain distance around a site,  at least until a feasibility
                                    study is completed. One county in New Jersey has done this. It ap-
                                    pears, however,  that  most  local  communities are  ignoring  or
                                    avoiding such decisions primarily because of a lack of information
                                    which would legally back up such controversial moves. To avoid
                                    the Love Canals or Times Beaches of the future, this issue needs to
                                    be addressed generically at the Federal or State level rather than on
                                    an ad hoc basis only.
                                    AFTERTHOUGHTS

                                      Public involvement in hazardous waste site decisions will occur in
                                    one form or another. It makes sense  that such involvement occur
                                    early with a goal of informed public input and, ultimately, consent.
                                    This ounce of prevention should be seen as a pragmatic step even
                                    by those who are impatient with public participation. States should
                                    aim for a program which is honest and direct and which may, as an
                                    added benefit, encourage an educated public to participate in pro-
                                    tecting the environment. Hopefully, hazardous waste site cleanups
                                    will not last forever, but there will probably always be hazardous
                                    waste to: manage, and engineers will always be looking for ways to
                                    do that as effectively  as possible. Achieving these goals involves
                                    everyone.

                                    REFERENCES

                                    1. Singer, G.L., "Nor Any  Drops to Drink!: Public Policies  Toward
                                       Chemical Contamination  of  Drinking  Water,"  (Princeton,  NJ:
                                       Princeton University, Center for Energy  and Environmental  Studies,
                                       PU/CEES ® 140) June 1982, Appendix C-l.
                                    2. New Jersey Department of Environmental Protection, Management
                                       Plan 1983-1986 For Hazardous Waste Site Cleanups in New Jersey,
                                       Trenton, NJ, Aug., 1983.
                                    3. New Jersey Department of Environmental Protection, Response Docu-
                                       ment for Public Hearing Comments on Management Plan 1983-1986
                                      for  Hazardous  Waste Site Cleanups in  New  Jersey,  Trenton, NJ,
                                       June 1984.
                                    4. The New Jersey Spill Compensation and Control Act was promulgated
                                       in 1977 prohibiting the discharge of petroleum products  and other
                                       hazardous substances, and requiring  that all  accidental spills be re-
                                       ported to the New Jersey Department  of Environmental Protection. It
                                       also imposed a tax on certain transfers of these materials and estab-
                                       lished a fund for the expeditious cleanup and removal of spills and the
                                       payment of damages.
                                    5. New Jersey  Department  of Environmental  Protection,  Status  of
                                       Hazardous Waste Site Mitigation in New Jersey, Trenton, NJ,  June
                                       1984.
                                                                                           PUBLIC PARTICIPATION
                                                                                               381

-------
         ELECTRIC REACTOR  FOR THE  DETOXIFICATION
                     OF HAZARDOUS  CHEMICAL  WASTES

                                          W.R. SCHOFIELD, Ph.D.
                                                    J.  BOYD
                                              D. DERRINGTON
                                           J.M. Huber Corporation
                                           Huber Technology Group
                                                 Borger, Texas
                                                 D.S. LEWIS
                                              Radian  Corporation
                                          Organic Chemistry Division
                                                 Austin, Texas
 INTRODUCTION

  The Huber Technology Group (HTG) of the J.M. Huber Cor-
 poration has developed a high-temperature pyrolysis process called
 an advanced electric reactor (AER) for the permanent detoxifica-
 tion of hazardous wastes. Two  series of tests  were conducted to
 demonstrate the effectiveness of the AER for treating soils con-
 taminated with hazardous wastes.
  Tests using polychlorinated biphenyls (PCBs) were conducted in
 September,  1983 to seek USEPA certification under the auspices of
 the Toxic Substances Control Act  (TSCA). This certification was
 received in May, 1984. A second  test series using carbon tetrachlor-
 ide (CCLJ was conducted in May, 1984 to provide additional infor-
 mation for a Texas Department of Water Resources permit applica-
 tion in compliance with the Resource Conservation  and Recovery
 Act.

 PROCESS OVERVIEW

  The HTG process consists of an AER (Fig. 1), plus downstream
 process gas cleaning equipment. The AER employes a new technol-
 ogy which rapidly heats feed materials to temperatures in the range
 of 4000°F to 4500T with surface heating rates  of 105°F to 107°F/
 sec  using intense thermal radiation  in the near infrared range.
  The reactants are isolated from the interior reactor wall or core
 by a gaseous blanket formed on the inside core surface by flowing
 nitrogen radially inward through the porous core walls. The only
 feed streams to the reactor are the hazardous waste and nitrogen
 which is used as a blanket gas. Both solids and gases have reactor
 residence times of 0.5 to 1 sec at  4000°F to 4500°F.
  Carbon electrodes are heated electrically and  in turn heat the re-
 actor core to incandescence. Heat transfer is accomplished by ther-
 mal radiative coupling from the core to the feed materials. The soil
 contaminants are either vaporized prior to pyrolysis or pyrolyzed in
 place on the particle surface. Because destruction of the hazardous
 waste takes place  by  photon  bombardment-induced pyrolysis
 rather than oxidation, typical combustion products such as carbon
 monoxide, carbon dioxide and oxides of nitrogen are formed, if at
 all,  in insignificant concentrations.  The principal products of soil-
 borne chlorinated organic waste treatment using the Huber process
 are chlorine gas, elemental carbon principally in  the form of carbon
 black and a detoxified solid which is free-flowing and granular.
  A simplified process diagram of a reactor configured for research
purposes is shown in Figure 2. The  solid feed stream is introduced
at the top of the reactor by means of a metered screw feeder con-
necting the feed bin to the reactor.
  The solid feed is gravity fed through the reactor where pyrolysis
occurs. After leaving the reactor, the product gas and waste solids
                                                                                      coinc'«•«*•»• f«
                                                                           Figure 1
                                                                    Advanced Electric Reactor
                                                  pass through two post-reactor treatment zones (PRTZs). The first
                                                  PRTZ  is an insulated  vessel  which  provides additional  high-
                                                  temperature (approximately 2000 °F) residence times of 0.5 to 1 sec
                                                  for solids and 5 to 10 sec for gases. Solid and gas-phase resident
                                                  times for both the reactor and the PRTZ can be independently
                                                  varied to achieve essentially any desired destruction efficiency. The
                                                  second  PRTZ is water cooled and provides additional residence
                                                  time (approximately 10 sec). However, its primary function is to
                                                  cool the molten soil particles below their fusion temperature to
                                                  avoid coagulation in the treated solid waste bin and to cool the gas
                                                  prior to downstream particulate cleanup.
                                                    Most of the detoxified, solid material exiting the second PRTZ is
                                                  collected in a solids bin which is sealed to the atmosphere. The ma-
                                                  jority of fine solids which do not settle out in the bin are removed in
                                                  the cyclone. The off-gas is then forced by a fan to a bag filter for
                                                  removal of any remaining fine particulate matter.
382
ALTERNATIVE TECHNOLOGY

-------
                                                                                           Table 1
                                                                                 Range of Operating Parameters
                          Figure 2
            Process Configuration for the Test Series
  The process gas cleaning train includes an aqueous caustic scrub-
ber for acid gas, or in this case chlorine, and typically two banks in
series of five parallel activated carbon beds, for removal of any
trace amounts of residual chlorine and organics. The cleaned off-
gas (composed almost entirely of nitrogen and water vapor) is then
emitted to the atmosphere. The low process gas flow rate, about
600 ftVmin for the commercial-sized unit, economically allows the
degree  of cleanup described here.
OPERATING PARAMETERS
  The PCB trial burn consisted of four tests conducted over 3 days.
In all cases, Aroclor  1260 was mixed with sand and a small amount
of carbon black to form a solid waste feed containing approximate-
ly 3000 ppm (0.3%)  PCBs.
  The  CC14 test series consisted of 17 tests conducted over 4 days.
The  "waste"  material for the  CC14  was composed of screened,
dried soil (less than 35-mesh), activated carbon and CC14. Carbon
tetrachloride was chosen as a surrogate based on its commercial
availability and refractory properties (i.e., it is difficult to destroy
by thermal means, resulting in  its high position on the USEPA's
hierarchy of incinerability).
  In pretest screen procedures, the volatility of CC14 proved dif-
ficult to overcome on a simple soil matrix. The vapor loss from the
feed  material was unacceptably high. Therefore, activated carbon
was added to significantly reduce the effective vapor pressure of the
CC14 so that relatively high concentrations could be tested in a solid
matrix.
  The  process  operating parameters for the two sets of tests are
shown in Table 1. The parameters for the PCB trial burn were held
essentially constant, while the parameters for the CC14 test series
were varied over a wide range of conditions. The actual conditions
for each CC14 test are shown in Table 2.

TEST RESULTS

  Sampling, analyses and data interpretation  for the PCB tests
were performed by Radian Corporation (Austin, TX). Radian also
provided analytical support to confirm Huber results for the CC14
tests. In each case,  the principal organic waste (POW) material
(PCB or CC14) was  sampled in the feed stream and the cyclone
outlet.  PCBs were also sampled at the stack while CC14 was sam-
pled  at the outlet to the charcoal beds just prior to the stack. Solid
samples were taken at the solid waste bin and baghouse in both sets
of tests, and the caustic scrubber and charcoal beds were sampled
during  the PCB  tests. Tests were also conducted to determined
fixed gas concentrations at the stack during both sets of tests.
  Polychlorinated dibenzodioxins (PCDDs) and furans (PCDFs),
volatile products of pyrolysis, HC1 and NOx were analyzed for the
PCB tests.
Parameters
Waste Concentration
Solid Feed Stack (I)
Waste Concentration -
Liquid Feed Stock U)
Solid Feed Rate (Ib/min)
Liquid Feed Rate (Ib/min)
Reactor Temperature (°F)
Nitrogen Feed Rate (ft 3/min)
NA = Not Applicable
1 Standard conditions = 520° R and 1 atm.
PCB Tests
0.3
NA
15.5-15.8
NA
4100
145

CCli, Tests
0.37-13.76
99
1.1 -40. a
3.3
3746-4418
104.3-189.7

Table 2
CC14 Test Program
Run Test Reactor
No. No. Temp "F
I 1 3803
2 4 4102
3 5 4104
4 7 3814
5 6 4385
6 3 3799
7 2 3813
8 8 3772
9 15 3808
12 14 3792
10 11 4091
11 9 3746
13 12 4118
14 13 4418
15 10 3782
16 16 3800
17 17 3800
N2 Flow
(ft 3/min)
184,1
189.7
185.5
104.3
184.1
184.7
184.3
189.3
188.8
189.3
189.2
189.3
189.6
189.4
189.1
190.0
190.0
Feedrate Cone.
(Ib/mln) (ICC1,,)
32.8 1.37
22.2 1.37
5.5 1.37
5.5 1.37
5.7 1.37
6.7 1.37
21.9 1.37
40.8 0.37
3.1 13.76
1.1 13.76
22.9 0.37
21.5 0.37
4.6 0.37
5.1 0.37
4.5 0.37
3.3 99'
3.3 99'
' 99% as reported by Vulcan Chemical Company
  Samples  were  taken using USEPA and  NIOSH methods,
sometimes with appropriate modifications. Analyses for the POWs
were carried out using electron capture gas chromatography (GC-
EC) for CC14 and capillary gas chromatography-mass spectrometry
(GC-MS) for PCBs, PCDDs and PCDFs.
  The principal objective of both tests was to determine destruc-
tion efficiencies (DEs) and destruction and removal efficiencies
(DREs) for the POWs. DEs were calculated using:
    DE =
Where:
    RF =
    RG =
    CF =
    GF =
                     RpCp -
                                          x 100%
                                              (1)
                           RFCF
feed rate (Ib/hr)
process gas rate at  cyclone outlet (Ib/hr)
POW concentration in the feed (mass fraction POW)
POW concentration in the process gas (mass
 fraction POW)
                                                                                   ALTERNATIVE TECHNOLOGY
                                                         383

-------
while DREs were calculated using:
    ORE =
                       RFCF
                                          x 100%
                                                          (2)
                            RFCF
Where:
    RS =
stack  gas rate or process gas rate at a point down-
 stream of the gas cleaning train (Ib/hr)
POW concentration  in the stack gas (mass  fraction
 POW)
  When feed concentrations based on both weighed proportions at
the time of mixing and analytical data were available, worst case
values are presented.

Results for PCBs

  PCB concentrations  in the process  and emission streams are
shown in Table 3. Results of PCB analyses of the potential wastes
from the process, treated sand, baghouse solids and scrubber liquid
indicate that  for PCBs these streams may be considered nonhazar-
dous for waste disposal. All results were well below 50 ppm, the
lower limit set by TSCA for hazardous wastes containing PCBs (40
CFR 761.60,  Subpart D). Results ranged from 0.5 ppb to 1 ppb for
treated sand, 24 ppb to 530 ppb for the baghouse filter catch and
0.29 ppb to  2.7 jig/1  for the scrubber  liquid. Results  for the ac-
tivated carbon indicated a concentration of 1 ppb for the charcoal
bed inlet post-test sample. However, this result is inconclusive due
to the possibility of inordinately high surrogate recovery in the ex-
traction procedure.
  Worst case DEs and DREs for PCBs are given in Table 4. DEs
range from 99.9995 to 99.99995%.  A  statistical analysis of the
destruction efficiency data employing a  Monte Carlo simulation in-
dicates that the  ranges of DEs for. the  reactor at 95% confidence
limits are 99.9995 to 99.999989%. These results indicate that the
TSCA requirement of 99.9999% can be  closely approached with no
downstream cleanup.
                           Table 3
   PCB Concentrations for Process and Emission Streams by GC-MS


                           Tent  1    Teal 2   Tent 3   Tent 4
                           9-27-83   9-28-83  9-29-83  9-29-83
Gas Streams ( LŁ/SCM)
Stack
Cyclone Outlet

0.23
4.1

0.03
4.1

O.lU
21

0.30
2.4
 Liquid Streams (pg/L)

    Scrubber Liquid          0.29      O.B6     0.76     2.7

    Pretest Scrubber Liquid  <0.14       NS     <0.14    <0.22

    Scrubber Feed Water     <0.14       NS       NS       NS
Solid Streams (US/R)
Feed
Treated Feed
Baghouse Filter Catch
Charcoal Bed (pg/g)
Pretest Charcoal (pg/g)
2530
0.0005
0.024
NS
0.003
3100
<0.0005
0.29
NS
NS
2710°
0.0006
NS
NS
NS
2710
0.001
0.53b
o.ooi r
NS
Standard Conditions = 528 °R, 1 aim.
NS = Nol Sampled
Values are not blank subtracted.
a. Single feed batch used for Tests 3 and 4
b. Cumulative for Tests 3 and 4
c. Cumulative for Tests 1 and 4
                                                                                 Table 4
                                                                  Destruction Efficiencies and Destruction and
                                                                       Removal Efficiencies for PCBs
                                                                  Pummeleni
                                                                                  Te»l 1
                                                                                           Tal 1
                                                                                                     Test 3
                                                                                                               Test 4
Destruction
 Efficiency ("To)    99.99990   99.99992   99.9995    99.99995

Destruction and
 Removal
 Efficiency (%)    99.999995   99.9999994  99.999998   99.999993
                                                        DREs differ from DEs; the product gas has passed through a
                                                     baghouse, caustic scrubber and charcoal beds prior to entering the
                                                     stack.  In all cases, the DREs easily exceeded the 99.9999% TSCA
                                                     criterion for PCB incineration (40 CFR 761.70 Subpart D). DREs
                                                     ranged from  99.999993 to 99.9999994%. A Monte Carlo simula-
                                                     tion of the DRE data for the process indicates that at 95% con-
                                                     fidence levels, the expected efficiencies range from 99.999991 to
                                                     99.9999997%.
                                                        PCDDs and PCDFs were analyzed in cyclone outlet samples us-
                                                     ing selected ion monitoring (SIM) GC-MS. The  results  of these
                                                     analyses show that the PCDDs and PCDFs were below the detec-
                                                     tion limits  of  from 0.03 /ig/SCM to 0 .06 Mg/SCM. For volatile
                                                     organics, toluene at a maximum concentration of 180 jig/SCM (45
                                                     ppb) and three  unknown hydrocarbons  at  lower  concentrations
                                                     were detected in 3 out of 10 stack samples. No volatile halogenated
                                                     organics were detected in the stack gases at detection limits of ap-
                                                     proximately 1  ppbv to 20 ppbv.
                                                        Particle loading analysis yielded a maximum loading of less than
                                                     7.1 mg/SCM at the stack. This is well below the 180 mg/SCM New
                                                     Source Performance Standard for incinerators (40 CFR 60,  Sub-
                                                     part E). Because of low concentrations, NOx data showed con-
                                                     siderable scatter. However, the highest average NOx concentration
                                                     was 16 mg/SCM (8.8E- 3 Ib/hr). No  applicable NOx standard for
                                                     incinerators is available for comparison. Chloride analyses for the
                                                     stack emissions yielded concentrations that were below detection
                                                     limits in all cases with a maximum concentration of less than 0.016
                                                     mg/SCM (^8.8E-6 Ib/hr). The  maximum chloride emission rate
                                                     was well below the criterion of  4  Ib/hr  for hazardous waste in-
                                                     cinerators ( 40 CFR 264). Fixed gas analyses of the stack gases by
                                                     Orsat indicated a minimum of 96.8% N2 with typical concentra-
                                                     tions greater than 99.5% and CO2 concentrations less than 0.2%
                                                     (detection limit) in all cases.
                                                     Results for CC14

                                                       On May 18, 21, 22 and 23, 1984, CCL,, a highly refractory hazar-
                                                     dous waste  surrogate, was used as a contaminant on a solid
                                                     substrate. Soil with activated carbon added was used in 13 tests, ac-
                                                     tivated carbon alone in two tests and a commercially pure liquid in
                                                     two tests.
                                                       The  four  primary  test variables (Table  2)  were:  (1) reactor
                                                     temperature (three levels—nominally 3800°, 4100 "and 4400 °F); (2)
                                                     feed rate (four levels—nominally 5, 20, 33 and 40 Ib/min); (3) con-
                                                     centration of CCU (four  levels—nominally 0.5,  2, 20 and 100%);
                                                     and  (4)  residence time—a  function of N2 flow and temperature
                                                     (two levels—nominally 3 and 6 sec).
                                                       The basic  Huber and Radian analytical data are found in table 5.
                                                     Tests 16 and 17 were conducted with CCL4 as a pure liquid feed.
                                                     The relatively high quantity of CC1< at the cyclone outlet indicates a
                                                     lower DE than with the soil-based feed. However, CC^ levels at the
                                                     stack inlet are almost indistinguishable from those of other soil-
                                                     based tests.  This clearly demonstrates the intrinsic  safety features
                                                     of the AER  to handle process upsets and incomplete destruction if
                                                     and when they occur.
                                                       Fixed gas  data (5 tests) indicated greater than 98.9% N2 in all
                                                     cases, with O2 ranging from 0.3 to 0.8% and  CO2 from undetected
                                                     to 0.6%.
384
          ALTERNATIVE TECHNOLOGY

-------
                           Table 5
          CCI4 Concentrations for Various Process Streams
                            Table 6
        Destruction Efficiencies (DE) and Destruction Removal
                   Efficiencies (ORE) for CCI4
Test Feed
No. (2)
1 1.24
2 1.14
3 1.34
4 1.75
s —
6
7 *--•-
8 0.28
9 0.46
10
11 —
12
13
14
15 13.76
16
17 —
1
Treated
Material
(ug/g)
0.69
0.47
—
0.47
—
—
0.07
0.14
—
0.56
—
—
—
—
0.18
—
—
Bag House
Filter
Catch
(ug/g)
_
—
—
—
—
--
24.10
4.20
—
1.33
—
—
—
—
—
—
—
Cyclone Outlet
(ug/scm)
Huber Radian
0.03 0.02
0.008
0.006
0.15 0.12
0.05 —
0.01
0.06
0.008 0.008
0.004 0.004
0.0008
0.003 —
0.02
<0.002 —
0.002
0.02
3300
5700 —
Cyclone
(ug/i
Huber
0.002
0.002
*
0.007
<0.0002
0.0007
0.0004
0.002
<0.0009
<0.0004
<0.0004
<0.0004
<0.0004
<0.0008
<0.0008
0.005
0.003
Outlet
jcm)
Radian
0.0007
—
—
0.0004
—
—
—
0.001
<0.0007
0.0007
—
—
<0.0004
—
—
—
—
 *The sampling pump failed while taking this sample.
 -- = not sampled.

  DEs and DREs for  CC14 are shown in Table 6. DEs  were
 99.9999% or greater in  most cases and 99.999% or greater in all
 cases except Tests 12, 16 and 17. The Radian crosscheck results
 supported these conclusions.
  The DRE results clearly  demonstrate the extremely high treat-
 ment capabilities of the HTG AER process. No test yielded results
 below 99.9999% DRE. This is over two orders of magnitude better
 than  RCRA minimum requirements  for  hazardous  waste  in-
 cinerators. Again, agreement between DREs  calculated  from the
 Huber results are supported by the Radian  crosschecks,  and both
 support the previous statement of high AER treatment capability.
AER APPLICATIONS
  Numerous applications exist for the AER including: hazardous
organic waste treatment; metal refining operations; production of
powdered refractories, fillers, glass fibers and other ceramics; con-
version of wastes to syngas; and vitrification of inorganic hazar-
dous wastes and low level nuclear wastes. In hazardous waste treat-
ment, this process is uniquely suited for:
•Treatment of low Btu content hazardous materials; i.e., contam-
 inated soils, pure PCBs and other heavily halogenated hydro-
 carbons
Test No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
DE
(*)
99.999920
99.999956
99.999908
99.999240
99.999018
99.999749
99.999296
99.999914
99.999922
>99. 999924
99.999952
99.998355
>99. 999873
99.999978
99.999916
98.3
97.1
ORE
U)
99.999992
99.999990
*
99.999963
>99. 999996
99.999987
99.999996
99.999981
>99. 999983
>99. 999961
>99. 999992
>99. 999961
>99. 999965
>99. 999991
>99. 999997
99.999997
99.999999
                                                                  •Sampling pump failed.
•Treatment of extremely hazardous materials; i.e., dioxins, PCBs
 and nerve gas
•"In-process" treatment of hazardous by-products from chemical,
 metallurgical and ceramical processes, frequently with profitable
 raw material recovery
  The  process offers a number of unique advantages including:
transportability; noncontact reactor design; extremely high process
temperature with relatively long residence times which result in very
high treatment  efficiencies; essentially no stack  or  fugitive emis-
sions; intrinsic safety features; and the ability to detoxify wastes in
a pyrolytic atmosphere,  thereby  avoiding  products of oxidation
such as dioxins and furans.
  HTG plans to use its Borger, Texas, 12 in. pilot-scale AER in a
number of research activities including: establishing DEs and DREs
and treatment costs as a function of operating conditions and pro-
cess  configuration; developing practical feed pretreatment pro-
cesses;  process development for various "in-process" applications;
and treatment feasibility testing with specific wastes from potential
treatment sites  (i.e., Love Canal,  S-Area and  other Superfund
sites). A commercial transportable unit with a throughput of 20,000
to 30,000  tons/year is being  designed for on-site soils detox-
ification  work.   It is scheduled to begin commercial operation in
mid-1985. Larger units are also being considered depending on  de-
mand.
                                                                                    ALTERNATIVE TECHNOLOGY
                                                          385

-------
                THE ECONOMICS  OF GROUND FREEZING
                FOR MANAGEMENT OF  UNCONTROLLED
                               HAZARDOUS WASTE  SITES

                                         JOHN M. SULLIVAN, JR.
                                        DANIEL R.  LYNCH,  Ph.D.
                                        Thayer School of Engineering
                                              Dartmouth  College
                                          Hanover, New  Hampshire
                                    ISKANDAR K.  ISKANDAR, Ph.D.
               U.S. Army Cold Regions  Research  and Engineering Laboratory (CRREL)
                                          Hanover, New  Hampshire
 INTRODUCTION

  Artificial ground freezing is not a new technology. There exists a
 100-year tradition of shaft sinking in which ground freezing has
 been used. The increasing application of ground freezing for civil
 engineering projects in recent years is mainly due to the following
 advantages:'
 •In principle, ground freezing can be used in all types of soils
 •Ground freezing is a  very flexible construction  method which
 can meet many boundary conditions and requirements
 •Very little or  no environmental  concern is associated with the
 method when dealing with soils for  civil  engineering purposes
  During ground freezing, the temperature of the soil water is
 lowered below the freezing point. The freezing temperature of soil
 solutions is not 32 T as for pure water, since dissolved ions in the
 soil lower the freezing  point.  However,  empirical relations exist
 that quantify the freezing point of soils.2'5 It might be argued that
 the freezing point of hazardous waste is much lower than that of
 soil systems. While this is a valid point, artificial freezing is done in
 the soil surrounding the hazardous waste and not in the waste itself.
 Therefore, uncontaminated  soil data are usable.  When the  soil
 temperature is lowered  to the freezing point, important changes
 begin to occur in soil properties. The strength of the soil is substan-
 tially increased and the soil permeability is  decreased. The potential
 use of ground freezing in hazardous waste remedial action is based
 on these two important points. The increase in soil strength upon
 freezing means that a frozen zone of  soil can be formed around or
 underneath a hazardous waste site or between the site and an un-
 contaminated environment without adding concrete, slurry walls,
 steel sheet pile walls or grout for injection.  Also, the frozen zone of
 soil becomes practically impermeable.
  The first use of artificial freezing was in  1862 in Swansea, Wales.
 The purpose was to support a mine shaft project used for mine pro-
 duction, material and personnel access, ventilation and emergency
 escape exits.
  In 1883 Poetsch patented a method of ground freezing with cool-
 ing pipes'  which, with some  modification,  is still  in use. In  this
 method, vertical drill holes with standard steel casings are uniform-
 ly spaced along the desired freezing line. Bore holes accommodate 3
 to 6 in. diameter pipe. Standard black pipe half the bore diameter is
 inserted in each casing, forming two concentric cylinders. A header
 or manifold system provides coolant such as calcium chloride brine
 at  -4°F to the interior pipe, with the return line being the outer
 casing. The manifold system runs along the  freezing line to reduce
 thermal  losses.  A  self-contained refrigeration system  pumps
 coolant around  the freezing loop.
                                                     An open loop system using an expendable coolant such as liquid
                                                   nitrogen (LN^) has the advantage over brine freezing because it
                                                   achieves a much lower temperature (- 321 T) in a very short time.
                                                   Therefore, LN2 is useful in emergency cases where time is limited.
                                                   Also, the fast freezing of contaminated soil by LN2 will result in im-
                                                   mobilization of  chemicals, as the soil water (with contaminants)
                                                   will freeze in situ.' Brine freezing, on the other hand, has the ad-
                                                   vantage  of  freezing the soil  walls in  a  more  regular  shape.
                                                   Temperature measuring instrumentation is appropriately placed for
                                                   monitoring the progress of the freeze front. A schematic represen-
                                                   tation of the two freezing methods is shown in Figure 1.

                                                                                      EXPADSIOn VALVE
                                                                                    COmPRESSOR
                                                                             Figure 1
                                                                Soil Freezing Methods — a) brine; b) LN2
386
ALTERNATIVE TECHNOLOGY

-------
  According to Braun and Nash,' the use of ground freezing in the
mining industry has  advantages over conventional methods (de-
watering, grouting, slurry walls, caissons):
•It does not require extensive geological data
•It serves several temporary functions such as support of an ex-
 cavation, groundwater control and structural underpinning
•It is adaptable to practically any size, shape or depth
•Excavation can be kept unobstructed as no bracing or sheathing
 is usually required
•It does not disturb the groundwater quantity or quality
•It is environmentally acceptable, as no chemicals will be added,
 and there is less disturbance to the site
Through 1978, more than 200 deep mine shafts had been driven by
artificial soil freezing.9
  In addition to its use in the mining industry, ground freezing has
been used for construction of open excavations and deep unsup-
ported construction trenches. For example, it was used during the
construction of subways in Moscow and in Zurich. 10'u About 70 in-
clined tunnels and over 30 excavations were made by soil freezing.
The use of ground freezing in the Moscow project saved 700 tons of
metals and 650 yds3 of timber, and the project was completed 11 to
12  months  early.10'11 This  project was circular, with  a  130  ft
  In North America, artificial freezing has been used  since 1888.!
In 1959, it was necessary to enlarge a twin railroad tunnel in Mon-
treal. Construction problems arose because of the presence of a
plastic layer of clay in the soil and because the tunnel  was located
under the city and ran beneath  service pipelines  and two large
buildings. Artificial soil freezing was successfully utilized in this
project.12
   In 1964, liquid nitrogen (LN2) was used for artificial  soil freezing
in Argenteuil, France.  In this  project  a collector sewage pipe
housed in a tunnel  broke.  The  sewage flooded the  tunnel and
seeped to a nearby stream. The influx was stopped by circulating
LN2 through 25 freezing probes. A concrete wall was later con-
structed between the polluted area and the fresh water stream.13
   The economics of ground freezing as a means of hazardous waste
containment are discussed below. These cost analyses are based on
existing construction practices and proven freezing technologies.
The data needed  to calculate thermal  parameters required  for
technical and economic assessments of ground freezing are routine-
ly obtained during the geotechnical and hydrologic site examina-
tions. This site-specific information  is required to evaluate  the
technical feasibility of the containment alternatives.14
   The  thermal data  are  obtained primarily  from soil texture,
moisture content and temperature measurements. The specific heat
of soils depends primarily on the water content since the volumetric
heat-capacity ratio for water to most dry soils is about  5. The ther-
mal conductivity of coarse-grained soils is significantly larger than
that of fine-grained  soils.  Both  saturated  soil types  exhibit a
decrease in  thermal conductivity with  increasing water content.
   Moisture content measurements determine the latent-heat energy
requirements and establish whether or not the  soil is saturated. A
saturated soil system  is desirable for an impermeable frozen barrier
and is assumed throughout this analysis. Lunardini15 provides ex-
tensive data relating these site examination measurements to soil
thermal properties. As an example, Figure 2 displays thermal con-
ductivity  as a function  of moisture  content  for  a  fine-grained
saturated soil.
 ECONOMIC CONSIDERATIONS
   There are no additional site examination  costs for the ground
 freezing treatment, as opposed to alternate containment modes.
 Therefore,  the economics of the site-specific investigation (i.e.,
 geotechnical, hydrologic and lab  filter-cake permeability testing)
 are unchanged from the USEPA estimate of $20,000 - $80,000.'"
   Unit costs for most of the equipment required for ground freez-
 ing  are shown  in Table 1.  Equipment mobilization  involves
 transport of the boring rig,  refrigeration units, piping and  site-
                                                                                             FINE-GRAINED SATURATED SOILS
          2.5 _
       e  1.5
           .5
                                           FROZEN SOIL
                   UNFROZEN  SOIL
                       \
                      10
'  \ '
 20
'  I '
 30
'  \ '
 40
'  \
 50
                        MOISTURE CONTENT (I DRY WEIGHT)
                             Figure 2
        Thermal Conductivity as a Function of Moisture Content
          for a Fine-Grained Saturated Soil (Lunardini, 1981)
clearing equipment. The site preparation requirements for ground
freezing are relatively low. The barrier must be saturated with water
if the soil moisture content is inadequate. Land clearing is necessary
for equipment access along  the  freezing route. Excavation and
heavy duty land clearing are not usually required for ground freez-
ing.  Capital costs include drilling and pipe  system expenses. The
drill-hole steel casings are not recovered at the completion of the
project. However, the header system and interior cooling lines can
be rented on a monthly basis. Energy requirements involve rental
of the refrigeration units, electrical consumption and expendable
coolants if used.

                             Table 1
       Unit Costs for Ground Freezing Equipment and Supplies*
                                     Daily
                                     Output
                    Total Costs ($)
                    100/unit
  1. Mobilization"
    Dozer, drill rig, refrigeration unit
    over 100 miles add

  2. Clear wooded lot (trees ^10 in
      dia.)"
    Grub stumps and remove
    Dozer medium duty clearing

  3. Header pipe system"
     70 GPM 3 in dia.
    150 GPM 4 in dia.
    400 GPM 6 in dia.

  4. Well hole drilling"
    4 in ID steel casing
    5 in ID steel casing
    6 in ID steel casing
    Drive shoe

  5. Black steel pipe"
    2 in dia."
    3 in dia.

  6. Self-contained refrigeration units"
      7 ton refrigeration
    110 ton refrigeration

  7. Liquid NI"
  8. Electricity
                    1/mile/unit
0.7 A +
1.5A
3,000V2
Month-
rental costs
per L.F. of
pipe
100 L.F.
2450/A
1100/A
0.31/Y2
1 2 3
1.40 0.85
1.60 0.90
2.50 1.00


0.65
0.70
0.75

                     9/L.F.
                    12/L.F.
                    15/L.F.
                    75/well


                    0.22/L.F./M
                    0.36/L.F./M


                     ISO/day
                    1.23/week

                    1.23/100  ft3

                    0.10 per kwh
 * All prices include parts, labor, operating and profit for subcontractor unless otherwise noted.
 " 2 in. pipe (S5.20/L.F.) — Rent at 2 yr. writeoff = 0.22/L.F./M
  3 in. pipe (S8.66/L.F.) — Rent at 2 yr. writeoff = 0.36/L.F./M
 + A = acre, Y2 = sq. yard, L.F. = lineal foot, M = Month
                                                                                        ALTERNATIVE TECHNOLOGY
                                                            387

-------
  The time constraint for the frozen wall plays a primary role in the
cost estimate. Mechanical refrigeration units rated ai 5-110 tons of
refrigeration  are  readily  available."  These units provide  the
manifold system with reusable coolant at  -4°F when operated
within  their  appropriate  capacity  range.  Expendable  LN2  is
available in large quantities when the demand for a rapid freezing
front is  required. For this system, the expanded N2 gas is vented
directly  to the atmosphere.  The refrigeration  units are replaced
with LN2 tanks and control valves that regulate the LN2 flow based
on the vent temperature.
   Sanger and Sayles™ provide a sound methodology for thermal
computations of frozen ground. Their energy  requirements and
freezing time estimates are somewhat more conservative than those
predicted  by finite element simulations and  actual  field
measurements.21'  u  However, for  this  preliminary  economic
analysis, their predictions are appropriate.  Sanger and Sayles
predict the expenditure of energy based on reasonable assumptions
about the  heat transfer  process in the soil.  The energy per unit
length, Q, time, t and power per unit length, P, required to freeze a
cylinder of radius R  is a function of the soil thermal properties,
thermal  conductivity, k,  thermal capacity, c, latent heat  of fusion,
L and the temperature difference between the coolant and soil."
   Ignoring second-order effects, they derived the energy estimate:
                                                           (1)
 where the first term in brackets accounts for the energy required to
 reduce the unfrozen soil temperature from T2 down to  freezing.
 The second term of Equation  1 is the energy associated with the
 transformation from unfrozen soil to frozen soil at the freezing
 temperature, i.e., the latent heat  of fusion,  L. The  last term
 describes the energy used in reducing the frozen  soil temperature
 from freezing to  the refrigeration temperature. The time required
 to freeze the column to a radius R is:
                                                                                      Table 2
                                                                            Symbol Definitions and Units
       R2L,
                                                           (2)
and the power requirement is:
             m(R/r0J
                                                           (3)
where the symbol definitions and units are as given in Table 2. The
total power requirement is larger than that expressed in Equation 3
due to inefficiencies in the refrigeration system. A 15% thermal loss
along the header system is  assumed. The refrigeration system is
conservatively rated at 0.21  tons of refrigeration per horsepower.'
The energy required for brine pumps and cooling fans is estimated
at 20% of the refrigeration load.
   The economics for ground freezing and slurry wall construction
are based on a 3 ft wall thickness. Once the soil columns merge ac-
cording to Equation 2, Sanger and Sayles approximate the frozen
soil thickness at 0.79  times  the soil column diameter. If this wall
thickness is less than 3 ft, the wall increases in  thickness as a planar
front according to separate equations published by  Sanger and
Sayles.10 This design thickness is a limitation of the slurry wall ex-
cavation equipment and  not  a result of structural  support or
permeability  requirements; nevertheless,  the authors have used it
for the frozen wall to establish a baseline comparison.
   The energy and time requirements are proportional to the square
of the radius of each cylinder (Eqs.  1 and 2). Initially, one might ex-
pect an economic advantage for a thin-wall construction via multi-
ple cylinders of small radius. However, the final cost analysis shows
                                                          ar     A factor which when multiplied by R defines the radius of tem-
                                                                 perature influence on the freeze pipe. Dimensionless—usually
                                                                 3 
-------
depth to the bedrock averages 40 ft. USEPA slurry wall estimates
and the authors' artificial ground freezing estimates for saturated
coarse quartz sand initially at 45 °F are shown in Table 3. The cost
has been plotted as a function of freezing rod spacing in Figure 3. It
can be seen in Table 3 that artificial ground freezing  is an accep-
table solution, provided the containment time requirement is short
(less than 135 days). Thereafter, the daily maintenance costs make
the ground freezing alternative unattractive.

                             Table 3
        Slurry Wall and Frozen Ground Construction Estimates
 Activity
                                     Unit Costs*   Total Costs
 A. SLURRY WALL"
 Testing—geotechnical, hydrologic and lab
  filter cake permeability                    N.A.         $ 20,000-$ 80,000
 Equipment Mobilization—hydraulic backhoe,
  bulldozer, slurry mixer, etc.                 N.A.         $ 20,000-$ 80,000
 Slurry trenching, excavation, mixing and
  backfilling                            $45-$70/Y2     $200,000-$310,000

 Maintenance
 Overall                               N.A.         $240,000-$470,000
 Average                              —           $355,000
 B. ARTIFICIAL GROUND FREEZING
 Testing—geotechnical, hydrologic and lab
  filter cake permeability
 Equipment Mobilization, clear, 4 in drill
  casing
 Rent—refrigeration, 4 in header, 2 in
  pipes, manpower
 Energy consumption
 Maintenance

 Extra melt time due to latent heat
  (numeric constant in Eq. 4 = .1614
 Overall**
 Average
   N.A.

   S21.4/Y2


   S6.9/Y2
   S5.7/Y2

   S0.31/Y2/
     day
$ 20,000-$ 80,000

$ 95,000

$ 30,500

$ 25,500
$1400/day
               25 days
Maintenance + $171,000-$231,000
        Maintenance + $200,000
•See Table 1 for unit costs. Y2 is sq. yards for depth x linear dimension. A 3-ft wall thickness is
   assumed in all calculations.
** Figure 3 at 18 day freeze time with 214 drill holes.
                                        Examining Figure 3, one can see that as the drill spacing becomes
                                     tighter, the fuel costs, equipment rentals and time for wall comple-
                                     tion are reduced.  These results  agree with Equations  1 and 2. A
                                     tight drill  spacing yields small frozen  soil column  radii. This
                                     reduces the overall energy requirement and permits use of less ex-
                                     pensive refrigeration equipment. The drawback of the close drill
                                     spacing is the expense associated with the drilling operation. The
                                     lineal footage of piping, a drive shoe for each well drilled and the
                                     labor charge per vertical foot drilled overwhelm all other economic
                                     parameters.

                                     Chemical Spill
                                        Consider the situation where a derailed chemical car disperses a
                                     toxic substance over an area adjoining a railroad track (Fig. 4). Sur-
                                     rounding towns impose  a time  constraint on the chemical and
                                     transportation  companies   for  containment  of  the  waste.  A
                                     preliminary week is required to define the hazardous spill and ob-
                                     tain general site test results.
                                                 CONNECTICUT  RIUER
                      FROZEN WALL 1000 X 3 X 40 FT
500 _



Ann
•fUU —
8 300 J
0
§
) —
g 200 J
100 J
o


SOLID - OVERALL COST
- DRILL EXPENSE
- FUEL COSTS
- EQUIPMENT RENTAL
- DAYS


V / ^"
\ .^^
^.^ ' _^^^
'<>- ^
.„.,_.,-., 	 •*-' 	
| , i i ,,,,,,,, | , . .
50

.
-
" in
— 4U
130 D
A
Y
S
120
110
• o

                       4          6           8
                          DRILL  SPACING  (FT)
                             Figure 3
               Economic Overview for EPA Test Case
                         (1000 x 40 x 3 ft)
                       10
                                                                                                    Figure 4
                                                                                  Hypothetical Train Car Toxic Spill — Plan View
                                                                                              TRAIN SPILL  130 FT RADIUS






o
g
g
g
CJ


250 _



200 _:

150 J
-
100 J
~
50 _:
n •


SOLID = OVERALL COST
- DRILL EXPENSE
- FUEL COSTS
	 - EQUIPMENT RENTAL
	 = DAYS

'
\ '' ^
K ' ^^
^^~~~~^ -L—^^^ •
' ^^ 	 f~ ~~ . •
1 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' T
_50



1-40

130
-
120
_
110
~ f\
L_ 0






D
Y
S



                                                   2345678
                                                               DRILL  SPACING  (FT)

                                                                   Figure 5
                                                Economic Overview for Train Car Spill Example
                                                                                         ALTERNATIVE TECHNOLOGY
                                                                                                  389

-------
    250 _
    2001
 §
 S
    100 J
     501
                COARSE-GRAINED, HIGH MOISTURE  SYSTEM
SOLID   -  OVERALL  COST
         -  DRILL  EXPENSE
	   -  FUEL COSTS
	     =  EQUIPMENT RENTAL
.._.._..-  DAYS
                                                                        FINE-GRAINED, HIGH MOISTURE SYSTEM
  50


140
                                           130  D
                                                 A
                                                 Y
                                                 S
                                          L20
                                           110
                               T
                            T
                    3      4     5      6     7
                        DRILL SPACING (FT)
250


-
2001

1501

-
-
1001
501
0 "

SOLID - OVERALL COST
"DRILL EXPENSE
~ - FUEL COSTS
7.7. • EQUIPMENT RENTAL
. ._. _. .- DAYS t
/
L /
\. ./
• ^-—L^^^^
\
' "-..;.
	 — ~— =
_50
.
-
-
140

130
.
-
•
120
110
' 0






D
A
Y
S



                                                   I
I
                                                                                 4567
                                                                               DRILL SPACING (FT)
                            Figure 6
      Economic Overview for Coarse-Grained, High Moisture Soils
                                                                                   Figure 8
                                                              Economic Overview for Fine-Grained, High Moisture Soils
                  COARSE-GRAINED, LOV MOISTURE SYSTEM
C Jtj — ,
_,
-
200J

150 -
_
-
-
100 J

_
501
0 "
SOLID = OVERALL COST
. . = DRILL EXPENSE '
= FUEL COSTS
7.7.~ = EQUIPMENT RENTAL '
. ._ _. .= DAYS
/

k
V^

f ^ —
^_^ -^"^
^ S
**"**" — • ~- • ' ' '

_ *r\l


L40

130 D
A
Y
S
120
.
_
110
' 0
             2345678
                        DRILL SPACING (FT)

                            Figure 7
      Economic Overview for Coarse-Grained, Low Moisture Soils
                                                                                    FINE-GRAUCa LOV MOISTURE SYSTEM
                                                                            SOLID    - OVERALL  COST
                                                                                     - DRILL EXPENSE
                                                                                     - FUEL COSTS
                                                                            7.7.7   - EQUIPMENT RENTAL
                                                                            	- DAYS
                                                                                                               50


                                                                                                             140
                                                                                                                       130  D
                                                                                                                             A
                                                                                                                             Y
                                                                                                                             S
                                                                                                                       H20
                                                                                                                       110
                                                                                            I
                                                                                        I
                                                                     2345678
                                                                               DRILL SPACING  (FT)

                                                                                  Figure 9
                                                              Economic Overview for Fine-Grained, Low Moisture Soils
  Initial drill samples estimate the barrier depth at 15 ft. Assuming
the pollutant diffuses horizontally I ft/day, the frozen wall is plan-
ned at a radius of 130 ft. This information is used to generate the
economic overview presented in Figure 5. The optimum cost design
calls for a 3.8 ft drill spacing with a 12-day freezing time. If there is
insufficient time remaining to freeze the soil before the time con-
straint is reached, the drill spacing is reduced with an associated in-
crease in overall costs.

Discussion
  The thermal properties used in  both of the above examples are
those determined by  O'Neill2' for saturated quartz sand. The fol-
lowing cases show the economic and time dependence as a function
of thermal parameters based on the train spill example geometry.
Using data from Lunardini"  for saturated soils the full range of
soil texture and moisture content effects is examined. The optimum
design configuration  for the various spills is given in Table 4 while
the economic overview of each soil system is plotted in Figures 6 to
9.
                                                         The results  show that increasing the soil moisture content in-
                                                       creases the time required to establish a frozen wall. For these high-
                                                       moisture soils, mechanical refrigeration would need a tight drill
                                                       spacing to satisfy the same time constraints as in the train car spill
                                                       case.  However, an expendable LN2 system with a 2.5-ft drill spac-
                                                       ing establishes an impermeable barrier within eight days of pump-
                                                       ing. This compares to a 22-day refrigeration time for a mechanical
                                                       system under  the  same conditions of  saturated fine-grained  soil
                                                       with a 40% moisture content (Fig. 8). The LN2 frozen wall assumed
                                                       a - 75 °F vent temperature for the freezing pipes.
                                                         The economics of expendable coolants are variable and generally
                                                       hard  to quantify. Veranneman  and Rebhan" approximate LN2
                                                       consumption at 800 kg of LN2  per m3 of frozen soil. Stoss  and
                                                       Valk" approximate  the LN2/brine expense ratio  at 2 for large
                                                       freezing projects (> 700m3) with  maintenance periods exceeding 30
                                                       days. Consequently, once the LN2 system establishes the barrier, a
                                                       mechanical refrigeration unit maintains the system during the waste
                                                       treatment process.
390    ALTERNATIVE, TECHNOLOGY

-------
                                                         Table 4
                                         Thermal Parameter Effects on Costs and Time
                                               Performances for Saturated Soils
Saturated
Soil
Texture

Coarse
Grain


Fine
Grain

O'Neill25
Thermal
'roper ties
ki
cal
cms3?
.00653


.00972
.00264


.00472

.009
cl cz
cal
cm3°C
.44


.44
.47


.46

.398
.71


.54
.72


.56

.589
L
cal
cm3
40


15
40


15

23.5
Moisture
Content
X of Dry
Weight
40


10
40


10


Cost
$/yard2 of
Perimeter
64


46
82


56


Time
Days
15


10
22


14


Figure
6


7
8


9

3,5
                       Cost estimate based on a wall  816 ft. round, 15 ft.  deep.

                       For comparison:  Slurry wall2"1      $75/y2 of wall
  An alternate economic overview is presented in  Figure  10 in
which a constraint on the maximum allowable  freezing time has
been introduced. The minimum cost for given geometric and ther-
mal conditions is  plotted as a function  of maximum allowable
freezing time; t* is the optimal (least cost) freezing time obtained
from the unconstrained Figure 5. If the time constraint is greater
than t*, then the optimum spacing is selected. For time constraints
less than t*, the cost rises following the curves as in Figure 5. Figure
10 was constructed using the train spill data.
                       TRAIN SPILL 816 X 15 FT
   o
   8
   140 J


   120 -


   100 -
t»


8   80 J


    60 J


    40.
                                                     _5
                                  MAXIMUM SPACING
                                                      _3
                                    MINIMUM COST
                   I ' '  '  ' I  '  ' ' '  I  ' '
                   5       10       15

                           TIME (DAYS)
                          Figure 10
            Cost and Drill Spacing as a Function of Time
25
      S
      P
      A
      C
 L-2  I
      N
      G
CONCLUSIONS
  Ground freezing as a means of hazardous waste containment can
be a cost effective operation for a large range of thermal condi-
tions. Soil parameters were shown to significantly affect the cost
analysis. Fine-grained soils with high moisture retention can double
the overall barrier expense compared to that of coarse-grained soils
with low moisture characteristics. However, regardless of the ther-
mal  conditions presented herein,  the  drilling operation was the
primary cost factor whenever a time constraint less than or equal to
the optimum spacing  was imposed. The economic  advantage of
ground freezing over  alternate barrier technologies is  limited to
temporary treatment sites due to the thermal maintenance expense.'

ACKNOWLEDGMENTS AND DISCLAIMER
   This work was supported by the USEPA, Disposal Branch, Solid
and Hazardous Waste Division, and the U.S. Army Cold Regions
Research and Engineering Laboratory (CRREL),  Hanover, N.H.,
under Interagency Agreement DW 930180-01-0. A portion of the
work was conducted at the Thayer School of Engineering at Dart-
mouth  College  in  Hanover.  The  authors acknowledge  the
assistance of Janet Houthoofd, project officer, USEPA. Citation
of brand names does not constitute an endorsement by either
USEPA or CRREL. The information contained in this publication
represents  the  authors'  opinions and not those of USEPA  or
CRREL.

REFERENCES
 1. Iskandar, I.K., Impact of Freezing on the Level of Contaminants in
    Uncontrolled Hazardous Waste Sites—Phase I, USEPA preliminary
   report, unpublished, 1984.
 2. Bouyoucos, G. J. and McCool, M.M., The Freezing Point Method as
   a New Means of Measuring the Concentration of the Soil Solution
   Directly in the Soil,  Michigan Agricultural Experiment Station Tech-
   nical Bulletin #24, 1915.
                                                                                   ALTERNATIVE TECHNOLOGY      391

-------
  3. Bouyoucos, G.J. and McCool, M.M., Further Studies on the Freezing
    Point Lowering of Soils, Michigan Agricultural Experiment Station
    Technical Bulletin #31,  1916.
  4. Bouyoucos, O.J. and McCool, M.M., "The Correct Explanation for
    the Heaving of Soils, Plants and Pavements," J. Am. Soc.. Agron-
    omy, 20, 1928, 480-491.
  5. Page, F.W. and Iskandar, I.K.,  Geochemistry of Subsea Permafrost
    at Prudhoe Bay, Alaska, CRREL SR 78-14, U.S. Army Cold Regions
    Research and Engineering Laboratory, Hanover, NH,  1978.
  6. ASHRAE Handbooks,  1981  Fundamentals  Handbook, ch.  1  and
    1982 Applications Handbook, ch 54, Pub. American Society of Heal-
    ing, Refrigeration and Air-Conditioning Engineers,  Inc.
  7. Iskandar, I.K.,  unpublished soil freeze data, CRREL,  1984.
  8. Braun,  B.  and Nash,  W.R., "Ground  Freezing Applications  in
    Underground Mining Construction," Proceedings, Third International
    Symposium on  Ground Freezing,  U.S. Army Cold Regions Research
    and Engineering Laboratory, Hanover, NH, 1982.
  9. Sadovsky, A. and Dorman, Ya.A., "Artificial Freezing and Cooling
    of Soils at  the Construction,"  Proc.,  Second International Sym-
    posium on Ground Freezing, Norway, 1980.
 10. Dorman, Ya.A.,  "Artificial Freezing of Soil in Subway Construc-
    tion," In Russian. Isd-vo. Transport, USSR,  1971.
 11. Tsytovich, el al., Physics. Physical Chemistry and Mechanics of Per-
    mafrost and  Ice, TL 439.  U.S.  Army Cold Regions Research and
    Engineering Laboratory, Hanover, NH, 1973.
 12. Low, G.J.,  "Soil  Freezing to Reconstruct a Railway Tunnel",  J.
    Const. Div., ASCE, 86., 1960.
 13. Stoss,  K. and Valk,  J., "Uses and Limitations of  Ground Freezing
    with Liquid Nitrogen", Engineering Geology, 13, 1979, 485-494.
 14. USEPA, Handbook for Remedial Action of Waste Disposal Sites:
    Final report  to OERR, ORD, MERL.  USEPA  Report No. EPA-
    62516-82-006, 1982.
                                                             15. Lunardini, V.J., Heat Transfer in Cold Climates, Van Nostrand Rein-
                                                                hold, New York. NY, 1981.
                                                             16. Geofreeze Subsurface Construction Co., Lorton, VA, personal com-
                                                                munication,  1984.
                                                             17. Means, Building Construction Cost Data—1982, 40 ed., Means, 1982.
                                                             18. King, P.A.  and Moselle, G., National Construction Estimator—
                                                                1984, 39 ed., Craftsman Book Co., 1984.
                                                             19. Merrimam-Graves Corp.—LN2 Suppliers, personal  communication,
                                                                1984.
                                                             20. Sanger, F.J. and Sayles,  F.H., "Thermal and Rheological Computa-
                                                                tions for Artifically Frozen Ground Construction", Ground Freezing,
                                                                FISGF, H. Jessberger, Ed., Elsevier Sci.  Pub., New York, NY, 1979.
                                                             21. Frivik, P.E., "State-of-the-Art  Report: Ground Freezing: Thermal
                                                                Properties, Modelling of Processes and Thermal Design", Ground
                                                                Freezing. SISGF, 1980.
                                                             22. Frivik, P.E. and Thorbergsen,  E.,  "Thermal  Design of  Artificial
                                                                Soil Freezing Systems", Ground Freezing, SISGF, 1980.
                                                             23. Carslaw,  H.S.  and Jaeger, J.C., Conduction of Heat in Solids, 2nd
                                                                ed., 287-288, Clarendon,  Oxford, 1959.
                                                             24. Spooner, P.A.,  Wetzel,  R.S. and Grube, W.  E.,  Jr., "Pollution
                                                                Migration Cut-Off Using Slurry Trench Construction", Proc. Na-
                                                                tional Conference on Management of Uncontrolled Hazardous Waste
                                                                Sites, Washington, D.C., Nov. 1982,  191-197.
                                                             25. O'Neill,  K.,  "Boundary Integral  Equation Solution  of Moving
                                                                Boundary Phase Change Problems", Int. J. Num. Meth. Engng.,
                                                                19, 1983, 1825-1850.
                                                             26. Veranneman, G.  and Rebhan,  D.,  "Ground  Consolidation  with
                                                                Liquid Nitrogen (LNi)",  Engineering Geology, 13, 1979, 473-484.
392
ALTERNATIVE TECHNOLOGY

-------
THE  ROLE OF ADSORPTION AND  BIODEGRADATION
               IN ON-SITE  LEACHATE  RENOVATION

                                   ROBERT C.  AHLERT, Ph.D.
                                         DAVID  S. KOSSON
                                       ERIK A. DIENEMANN
                                       FREDERICK D. RUDA
                               Chemical & Biochemical Engineering
                                          Rutgers University
                                       Piscataway,  New Jersey
 INTRODUCTION

   In  a number  of  recent papers, the authors have described
 experiments  with  high-strength,  hazardous  industrial
 wastewaters.1'2-3'4 These experiments have employed biodegrada-
 tion in beds packed with soil to oxidize organic solutes in waste li-
 quors. Laboratory and  pilot-scale bioreactor columns have been
 used; at both scales, vacuum has been applied at the base of vertical
 columns to balance capillary forces and mimic so-called "field
 status". The goal of these experiments is to develop design criteria
 for in situ microbial treatment immediately in or adjacent to an un-
 controlled dump or spill site.
   The approach  allows  natural selection to control the microbial
 community as completely as possible. External control is achieved
 through management of independent parameters such as soil type,
 depth to groundwater,  loading rates, nutrient additions,  etc. A
 mixed microbial population is established in the soil structure. The
 indigenous  microflora of the soil is supplemented through addi-
 tion,  at the soil  surface, of an inoculum of a mixed microbial
 population derived  from  the secondary sludge  of a municipal
 sewage treatment facility.
   The microbial  seed propagates  through  the  soil column and
 permeates the soil structure. Leachate feed is added at the soil sur-
 face and is allowed to diffuse through the soil, where it is subse-
 quently adsorbed and/or degraded through aerobic and anaerobic
 processes.
   An experimental field apparatus  was designed to examine this
 treatment process, (1) on a scale suitable for process design, (2)
 under natural environmental conditions and (3) over a prolonged
 period of time. To fulfill these goals, a pilot-scale treatment system
 was installed. The installation consists of six soil columns, termed
 self-contained lysimeters (SCLs), 60 cm in diameter and 120 cm
 deep. The SCLs were designed for complete effluent recovery, im-
 planted in the ground and operated in simulation of field  condi-
 tions. Data for bioreactor modeling was obtained from Experiment
 0682, started in June, 1982 and lasting 161 days.
   Numerous laboratory columns have been operated in sterile and
 mixed microbial modes, also. The sterile mode employs chloride
 ion as the market species and no nutrients. It is intended to examine
 residence time distributions based only on physical properties and
 hydraulic behavior. One of the goals of the overall research activity
 is the correlation of performance data for different soils and vary-
 ing reactor diameters to define generalized design criteria. For this
 reason, various laboratory packed bed studies have been run in
 parallel with the pilot plant experiments.
   The packed bed bioreactor is a complex system  that operates
 with convective and dispersive flow contributions, physical adsorp-
tion and chemisorption, catalyzed non-biochemical reactions and
aerobic and anaerobic mixed microbial reactor domains. With the
diversity  of hydraulic,  physical and  chemical influences,  the
development of performance correlations and generalized design
criteria is very  complex. Fully deterministic modeling is not pos-
sible. Thus, a step-by-step approach using incomplete models, as
appropriate to limit state operations, was chosen. Ultimately, it is
planned to couple these models in a comprehensive design scheme.
  Brief discussion of two limit cases follows. The first case is that
of dispersion with biochemical reaction. At some point after start-
up,  adsorptive capacities are saturated and output responses to in-
put  variations  are primarily functions  of empty  bed  flow rate,
dispersion and biochemical reaction. If the latter is taken as first-
order, a simple model results. This assumption is reasonable for a
mixed microbial system with many metabolic pathways. This case is
termed the end-state or semi-steady-state case.
  A second model has been associated with initial processes, i.e.,
startup. In this initial unsteady condition, biochemical reaction is
minimal and the packed bed reactor is viewed as an adsorber col-
umn. Macro-diffusivities from the first case can be applied in the
second case. This permits description of adsorption/desorption ef-
fects during the early stages of column operation.


DISPERSION MODEL

  A dispersion model in which macro-diffusion is superimposed on
plug-flow can be used to describe some reactor systems. This model
is frequently employed for flow through packed beds and is an ap-
propriate approach to bioreactive packed soil columns:
V.(EVc)   u.Vc t (i(c)
                                                      (1)
where 0 is the average fluid field velocity and (c) is a reaction rate
and/or source term. For an isothermal,, incompressible fluid at a
constant flow rate in a cylindrical vessel, Equation 1 can be written:
    .
  at
          E (
           r
                       2-
                                                      (2)
where Ez and Er are the axial and radial dispersion coefficients,
respectively. Both are assumed independent of concentration and
position. For plug-flow with negligible radial dispersion, Equation
2 reduces to:
  3t
                    s-ff"
                                             (3)
where r is the rate of reaction.
                                                                               ALTERNATIVE TECHNOLOGY
                                                      393

-------
   This  model  is  called the  "longitudinal-dispersion plug-flow
model" or, sometimes, simply the "dispersion model". Langmuir'
discussed this model and obtained the steady-state solution, using
the following boundary conditions:
                                                                           INFLUENT  ORGANIC  CARSON
          z=L
       -z-0
   For a pulse input, Equations 3 through 5 reduce to

    e(t)      >°°^A!;r->explL-L4"rYL    kti
                                                          (4)
                                                 (5)
                                                 (6)
   The effective flow area and average fluid velocity are obtained
 from calculated mean residence time and soil porosity, i.e.,
              x  e
            AXS  *
                 ,1000,
                  --
                                                 (7)
                                                 (8)
                                                 (9)
   Before proceeding with estimates for Ez and k (rate constant),
 mean residence times must be derived. For this purpose, statistical
 methods were employed.
   Statistsical comparisons of influent and effluent organic carbon
 concentrations were used to obtain values of mean residence time
 for use  in Equations 7 through 9. The method employed was the
 F-test in a one-way analysis of variance (ANOVA) for two popula-
 tions, each representing one level of treatment (influent and ef-
 fluent).
   Columns 1 and 2 from field experiment 0682 were examined in
 this study. The influent and effluent organic carbon concentrations
 for SCL 1 are shown in Figure 1. It was assumed that initial effluent
 peaks were a result of leachate adsorption coupled with microbial
 adaptation. For this reason, effluent data from days 1  to 75 were
 not used. Also, it was concluded that effluent peaks from day 75 on
 were a result of the influent peak between days 57 and 67.
   F-tests were conducted  over  a range from  20 to 60 days in
 lagtime, with influent data from  days 39 to 120. Actual F-testing
 was conducted with data produced  from 6-day  central-moving-
 averages. All influent and effluent data were reduced by 3  days at
 each end. Therefore, the normalized effluent data go only to day
 158.
   The influent and effluent organic carbon concentrations for SCL
 2 are shown in Figure 2. Again, the effluent data from days 1 to 75
 were deleted and  testing was confined to the remaining effluent
 data and was assumed to be in response to the  influent peaks be-
 tween days 57 and 67. F-tests were conducted for lag times between
 20 and 60 days with influent data from days 43 to 109. All points
 were  normalized  and developed  from  6-day  central-moving-
 averages.
   F-distributions as functions of lagtime  are shown in Figures  3
 and 4. The curve for SCL 1 peaks at a lag of 39 days. This signifies
 that 39 days are required before an influent perturbation is "seen"
 in the effluent  stream. In other terms, particles of mass spend an
 average  of 39 days in the column. Similarly, Figure 4  for SCL  2
 shows a  single peak, at day 50. Column dimensions and calculated
 values according to Equations 7 through 9 are given in  Table  1.
   Using  calculated values for porosity, average fluid velocity and
 effective flow area, a first approximation for Ez can be calculated.
 In Equation 6, k was first assumed equal to zero.  Figures 5 and  6
 describe  predicted concentrations versus time for SCLs  1 and 2 in
 addition to actual effluent concentration data. A value of Ez equal
 to 0.3  cmVhr appeared to be a reasonable first guess for both col-
 umns.
   At this point, estimation of the rate constant (k)  was possible.
 Returning to Equation 6, k was allowed to vary while Ez remained
                                                                                  TT'TTTT"1!	TT"TT
                                                               11 ii I? a n n it « n H S7 11 n n n a n a r 101 in in IB IB 111 i«t
                                                                                            DAY
                                                                                  Figure la
                                                                            Influent Organic Carbon
EFFLUENT ORGANIC CARBON
                   CVUIllfXI DM2
                   CO. I. «•*» IM
                     f Mvln^ rfwuf
                                                                                                     T-TT""
                                                               4 7 iz IT n ?7 n n « « s n H n 77 n o n n «? ::< m in in in m u« in
                                                                                            DAY
                                                                                  Figure Ib
                                                                            Effluent Organic Carbon
                                                                                   Table 1
                                                                    Pllol Packed Bed Properties & Performance
                                                                                            SCL1
                                                                                                           SCL 2
109.:
2858
312.2
40.6
1160
39
3.25
0.117
ir 0.187
0.00351
101.6
2858
290.4
40.6
1160
50
2.36
0.085
0.116
0.00188
                                                        Length, cm
                                                        Cross-section area, cm2
                                                        Volume, 1
                                                        Porosity, °7o
                                                        Effective flow area, cm:
                                                        Mean residence time, days
                                                        Ave. volumetric throughput, I/day
                                                        Ave. velocity, cm/hr
                                                        AxiaJ Dispersion coefficient, cm2/hr
                                                        Ave. rate constant
                                                        fixed at 0.3 cmVhr. It was found that k alters both the magnitude
                                                        and location,  or skew,  of the peak. For a first approximation,
                                                        magnitude was deemed most important. Thus, a value for k equal
                                                        to 0.0075 hr - was obtained for SCL 1. Similarly, k was approx-
                                                        imately 0.0015 hr -1  for SCL 2. Optimum values of Ej and k are
                                                        summarized in Table 1. In Figures 7 and 8, predicted data are com-
                                                        pared to observed data; agreement is good. Final values of E^ and k
                                                        were based on simultaneous regression of both Ez and k over nar-
                                                        rower  ranges of values to minimum  values of the objective func-
                                                        tion.
394
ALTERNATIVE TECHNOLOGY

-------
                    INFLUENT ORGANIC CARBON
                                         EXPCRIHENT OS82
                                         SCI. 2. Sandy lioi
                                         6 day «ov!m) ov«rag«
                                                                                               F DISTRIBUTION
                                 |impiTii|ira|rainn.iiii|	i|i«ii|i»|«mi|	i|miu|i
       < 7 12 17 22 27 32 17 42 47 S2 57 62 67 72 77 92 17 92 17 10* 111 111 125 132 139 146 IS3

                                         DAY
                            Figure 2a
                     Influent Organic Carbon
                                                                         so-


                                                                         40-


                                                                         30-


                                                                         20-


                                                                         10-
                                                                                    T 0682
                                                                              SO. 2. Sandy U-
                                                                              TOC
                                                                     i' i  T^ i' i'  r11' i  i   r i   i   i  i' I   i   i   i  i   i   i
                                                                    20 2:  24 26  28  30  32 34 36  3)  40  42  44  46  4B  SO  52  SI  56  58  60
                                                                                         LAG TIME  (days)

                                                                                              Figure 4
                                                                                       F Distribution, SCL 2
     110-
     100-
  (_> so-

  """ 40-

     30-

     20-

     10-
                     EFFLUENT  ORGANIC CARBON
                                  EWHIIKENt 0112
                                  Sa 2. S«idy Leo
                                  6 day Mivlng av«rag«
4712 1722273237424752576287727712879297 104 III 119 125 132 139 146 IS3
                               DAY


                     Figure 2b
              Effluent Organic Carbon
                    DISPERSION MODEL
                                     EFFIUCKT rod Concwitnttloni tuli
                                     Sa 1. Sandy Lou
                                     6 day Mvln] avwaga
                                                                                                                k=0 .ir"1
                                                                                                                      0.2
                                                                                                                              E, (cmVhr)
                                                                               lil'T'I'T'l'I'I'I'I'l"!"!11!"!"!"!"!"!"!"
                                                                              79 90 92 14 96 98 90 92 94 96 98  101  104 107  110  113 116  119 122  I2S
                                                                                                             DAY
                                                                                                      Figure 5
                                                                                              Dispersion Model, SCL 1
                        F DISTRIBUTION
   i.s-

   1.4-

   1.2-

   1.0-

 F .1-

   .1-

   .4-

   .2-
                                EXPERIMEHT 0692
                                Sa 1. Sandy Uai
                                TOG Concantratlan Bail*
     JTI' I' I '  I '  I ' I  ' I '  I '  I ' I  ' I  ' I '  I ' I  ' I  ' I '  I '  I ' I  ' I
     20  22 24  26  29 30 32  34  31 39 40  42  44  46 48  SO  52  94 58 51  60

                         LAG TIME  (days)


                            Figure 3
                      F Distribution, SCL  1
160-


140-
                                                                      40-


                                                                      20-
    DISPERSION MODEL
                     CfTLUCNT TOO Caneantratlan 8a.l«
                     Sa 2. Sandy Uoi
                     6 day Mvlng avtraat

k=0 hr"1
                                                                                                              0.2
                                                                                                              0.3  f  E, (cm2/hr)
                                                                        rrTi 111' i ]" i" i" i" i " i" r' r'  r' i" i" r' i  •
                                                                       94  98 99  101   104  107  110  113  116  119  122  125  129  131  134  137  140
                                                                                                          DAY
                                                                                               Figure 6
                                                                                       Dispersion Model, SCL 2
ADSORPTION/DESORPTION EFFECTS

  If r is replaced by OC ,  an expression for the relative rate of ad-
sorption/desorption, Equation 3 is relevant to a pure unsteady-
state adsorption column. This is the situation during treatment col-
umn startup; biochemical reaction is minimal because of the low
                                                                  density of microbial populations. Physical  and  non-biochemical
                                                                  surface binding dominate.
                                                                    Let a be given by:
                                                                                u ia
                                                                                n'at
                                                             (10)
                                                                                             ALTERNATIVE TECHNOLOGY       395

-------
where p is solid phase density, 17 is soil moisture content at satura-
tion and  q is concentration of solutes  on  solid surfaces (as
mass/mass). Further, assume simple partitioning of solutes
           k'c
                                                         00
In the absence of biochemical reaction, Equation 3 becomes
               32
or
      c
where

     R
ail
 n
                                                         (13)
                                                         (14)
 is a retardation factor. The boundary conditions of Equations 4
 and 5 are appropriate. The desired "initial" condition is a step, not
 a pulse,  in this case, i.e..
           0  at   al 1  x  and   t   «  0"

           c   at   x ^ 0  and   t > 0
                                                         (15)
      160-

      140-
                     EFFLUENT ORGANIC CARBON
                                       sa i.
                                       Mn-lln
              E    .1865  cmVhr
k   .3513x10-' nr
                   I     I
                  K    94
                               I
                              to:
                                  I
                                 104
                                        DAY
    140-

    120-

  — 100-
  1_
  Ol
  •.-  JO-

  IT
  -E  M-

  °40-

     20-
                          Figure 7
                Effluent Organic Carbon, SCL 1
    EFFLUENT ORGANIC CARBON
                       SO. I. la«lv U-
                       Ma*-lln«tf- ftigrtMlan

 EZ  '  .1350 cm'/hr

 k •  .1628x10''  hr"1
              1	1	1	1	1	1	1	1	1	1	1	1
             102   104   101   101   110   112   114   IK  III  120  122  124

                                        DAY
                          Figure 8
                Effluent Organic Carbon, SCL 1
                                                      uw

                                                      1000

                                                      2soo-
                                                                           015PERSION  MODEL
                                                                                 Without Biodegradatlon
                                                                                 Without Adsorption
                                                                                                           LC'l,  Exp.  0183
                                                              I—E2-5 cmVday
v»-
I
J-v
I 1
i it
15
1
13

1
20

I
a i

1 1
0 E
* **«»»»
1 1 1 1 i
1C 13 30 iS U
                                                                            TI-E  (days

                                                                             Figure 9
                                                                               Figure 9
                                                       Dispersion Model — Without Biodegradation; Without Adsorption
                                                                          DISPERSION MODEL
                                                                                 Without  Biodegradation
                                                                                 E2-1S c»2/day
                                                                                                        , LC'l .  Exp.  0183
                                                                                                                   .
                                                                                     IS   20   f!   »   B    10   ij   S3   S3   K
                                                                                            T!MŁ (days)

                                                                                            Figure 10
                                                                     Dispersion Model — Without Biodegradation; Ej = 15 cm2/day
  Equations 13 and IS were solved with numerical methodology,
i.e., the "classic explicit" finite difference methodology described
by Lapidus  and Finder.* Effluent concentrations were averaged
over one day intervals to mimic experimental sampling conditions.
  Sensitivity of the equations  to  variations  in Ez  and R^  is il-
lustrated in Figures 9 through 11. In these plot, the numerical solu-
tion is compared to experimental data obtained from a laboratory
column  (LC) during Experiment 0183. It can be seen that the solu-
tion  of  Equations 13 and  IS is  considerably more  sensitive to
changes in R(. than to changes in Ez. The best fit of experimental
data from LC 1 was obtained for R,. equal to 1.7 and Ez equal to IS
cm2/day or about 0.63 cmVhr.
  It was assumed that similar hydraulic and adsorptive conditions
prevailed for LCs 2 through 4, during Experiment 0183. These col-
umns were packed with the same soil and operated in like fashion
but received feed solutions of varying concentrations. The influent
TOCs for LCS 1, 2, 3 and 4 during the experiment were 3000,  2000,
1000  and 210  mgC/1, respectively.  The results  for the numerical
solution for these cases, assuming  R equal  to 1.7 and an Ez  of 15
cmVday, are presented in Figure 12. Experimental data are includ-
ed, also. The Ez estimated for the laboratory columns compares
favorably with the dispersion coefficients obtained from field data.
They  vary by less than a factor of five, i.e., order-of-magnitude the
same  for such soil properties.
  The model based on dispersion  and adsorption is a reasonable
approximation of the processes that occur  initially during startup
of the bioreactor columns (Fig. 12). This is an  expected response to
the relatively small microbial populations present at  the beginning
of an experiment. As a bioslime develops, removal of organic car-
3%       ALTERNATIVE TECHNOLOGY

-------
                       DISPERSION  MODEL
                             Without Biodegradation
                             R = 1.7 (Adsorption)
                                       A LC#1,  Exp.  0183
                      cm'/day
           I    I    I    I     I    I    I    I    I     I    I    I
      0    5   10   15    20    25   30   35   ID    45    50   55   GO
                          TIME  (days)

                          Figure 11
 Dispersion Model — Without Biodegradation; R = 1.7 (Adsorption)
                        DISPERSION  MODEL
                              R=1.7  (adsorption)
                                =    cnT/day
                                      LC#1,  Exp.  0183
                                    o LC#2,  Exp.  0183
                                    n LC#3,  Exp.  0183
                           TIME (days)

                          Figure 12
    Dispersion Model — R = 1.7 (Adsorption); Ez = 15 cmVday
bon by biodegradation becomes increasingly important. The dif-
ference between predicted and actual TOCs becomes significant
after 20,  10 and 7 days, respectively, for LCs 1, 2 and 3.
  With decreasing influent  TOCs, biodegradation becomes the
controlling process more rapidly. This result is most likely because
a larger fraction of the influent TOC is in the form of glucose for
the lower overall TOC levels. Thus, the organic carbon in the feed
streams to LCs  2 and 3 was more readily assimilated than that fed
to LC  1, and the acclimation period was shorter.
CONCLUSIONS
  A multi-faceted approach has been used for parameter estima-
tion and modeling of the complex, interacting processes occurring
in a soil-based, bioreactor system. Steady-state was assumed for
microbial and adsorptive processes after a prolonged period of
operation. This  assumption permitted the application of a modified
statistical  test  to  determine  packed bed  porositied  and  mean
residence times  for two pilot-scale columns. In turn, this informa-
tion was used in a dispersion model incorporating first-order reac-
 tion kinetics. Dispersion coefficients and pseudo-first-order rate
 constants were calculated for field experiments.
   Laboratory-scale columns were used to study the competition
 between physical  and chemical processes during  reactor column
 startup. A dispersion model with first-order reversible adsorption
 kinetics has been shown to be useful for the analysis of column per-
 formance before a significant bioslime develops. Modeling for the
 transition regime between the two limit cases is in progress.
   Step-by-step  simplified modeling has  been found  to  be  a
 reasonable initial step to the development of design criteria, scaling
factors and loading estimation for in situ leachate and spill liquor
 treatment.

 ACKNOWLEDGEMENT

   The authors with to express their  appreciation for the financial
 support of the Office of Water Policy (U.S. Department of the In-
 terior) and  the  U.S.  Geological Survey, through the New Jersey
 Water Resources Council.

 NOMENCLATURE
 A     effective flow area, cm2
 Axs    column cross-section area, cm2
 c      concentration, mg/liter
 Er     radial dispersion coefficient, cm2/hr
 Ez     axial dispersion coefficient, cm2/hr
 k      first-order reaction rate constant, hr -1
 k'     partition coefficient, 1/mg
 M     mass of pulse  input, mg
 Q     ave. volumetric throughput, I/day
r      reaction  rate, mg/1, hr
 Rc     retardation factor, dimensionless
t      time, hr
u      ave.  velocity, cm/hr
V     volumetric  capacity, 1
z      axial position (0 ^ z •Ł L), cm
a      adsorption/desorption rate, mg/1, hr
e      porosity, dimensionless
r)      moisture content  at saturation, dimensionless
p      solid phase density, mg/1
6      mean residence time, hr

REFERENCES
 1.  Kosson, D. and Ahlert, R.C., "In Situ and On-Site Biodegradation of
   Industrial  Landfill Leachate," Environmental Progress, in press.
2.  Kosson, D.  and  Ahlert, R.C., "Treatment of  Hazardous Landfill
   Leachates  Utilizing  In Situ Microbial Degradation," Proc. Manage-
   ment of Uncontrolled Hazardous Waste Sites, Washington, DC,  Nov.
   1983, 217-220.
3.  Kosson, D., Dienemann E. and Ahlert, R.C., "Treatment of  Haz-
   ardous Landfill Leachates Utilizing In Situ Microbial Degradation,
   Part II,"  Proc. Hazardous Wastes and Environmental Emergencies,
   Houston, TX, Mar. 1984, 289-292.
4.  Kosson, D.,  Dienemann, E. and Ahlert, R.C., "Characterization and
  Treatability Studies  of an Industrial Landfill Leachate (Kin-Buc I),"
  Proc. 39th Annual Purdue Industrial Waste Conference, W. Lafayette,
   IN, May 1984, in press.
5. Wen, C.Y. and Fan, L.T., Models for Flow Systems and Chemical
  Reactors, Marcel  Dekker, Inc., New York, NY, 1975.
6.  Lapidus, L. and Finder,  G.F., Numerical Solution of Partial Differen-
   tial Equations in Science and Engineering, Wiley, New York, NY,  1982.
                                                                                    ALTERNATIVE TECHNOLOGY
                                                          397

-------
THE USE  OF A  MICRODISPERSION  OF  AIR IN WATER  FOR
         IN  SITU TREATMENT  OF HAZARDOUS  ORGANICS

                                      DONALD L. MICHELSEN,  Ph.D.
                                          DAVID A.  WALLIS, Ph.D.
                                             FELIX SEBBA, Ph.D.
                                                   Virginia Tech
                                     Department of Chemical Engineering
                                               Blacksburg, Virginia
INTRODUCTION

  In this paper, the authors describe the use of a microdispersion
of air in water for treating hazardous organics. In particular, the air
microdispersion generated and tested in this study was a 60 to 65%
mixture of air  bubbles, 25 to 50 /»  in diameter and water, often
referred to as Colloidal Gas Aphrons (COAs).
  In these microdispersions, air  bubbles are encapsulated in thin
"soap" films that are so tenacious that the bubbles do not coalesce,
even when pressed together. Hence, the bubbles remain very small
(on the order of 25 /i) and present an enormous surface area. One
litre of CGA containing 60% of  gas as 24 n diameter bubbles con
tains 8.3 x  1010 bubbles with a surface area of 150 m^, about 1/6
the  area of  an  Olympic swimming pool. These air dispersions or
CGAs can be generated using virtually any water soluble surfactant
with concentrations as low as 100 mg/1 and any gas of limited
solubility (e.g., nitrogen, oxygen, carbon dioxide). With selected
surfactants,  dispersions can be generated containing 65% gas by
volume. CGAs must be clearly  distinguished from  the so-called
"bubbles" produced by dissolved air precipitation, sparging or
electrolysis,  all  of which are 2 to 1000 times larger and rise to the
surface rapidly where they quickly coalesce. Recent papers by
Auten and Sebba1 and Sebba,2'3 provide a good background on the
characteristics of CGAs.
  Because these fine microdispersions are so stable and small, they
can remain suspended in solution and can  flow through channels
such as exist in a sand bed. Larger, unstable "bubbles" are filtered
out in  such a situation.  Another important characteristic is that
CGAs  can be pumped with many  positive displacement pumps
without deterioration.
  A description of CGA production has been given by Sebba and
Barnett.4 "CGAs can be made very simply by passing a very rapid
stream  of dilute (about 2 x  10~3M) surfactant solution through a
venturi throat, at  which point there is a  very restricted orifice
through which gas (usually air) under an excess pressure of about 1
atm is sucked into the stream. Because of the turbulent jet and the
slow entry of gas, it enters in the form of microbubbles. A require-
ment for the generation of a shell encapsulated bubble is that the
gas  break through an aqueous-gas surface, with  a surfactant
monolayer at the surface, at least twice. The turbulence ensures
that. In contrast, if a so-called "bubble" is introduced by sparging
through a fitted disc or by gas precipitation, this is likely to be gas
surrounded  by the bulk water and  have only one interface, and
therefore, is effectively a gas-filled hole. To obtain a high concen-
tration of CGA bubbles, the suspension is recycled a few times
through the CGA generator. The method described for generating
CGAs is excellent for laboratory production, and a generator of
this type can operate for many years."
                                                      In the laboratory, CGAs have since been generated using a spin-
                                                    ning disk in a baffled chamber. The production of CGAs can also
                                                    be achieved using a high speed  commercial blender. These pro-
                                                    cedures have been scaled-up with  effective  production in a  19 1
                                                    stainless steel continuous (stirred) tank. More recently, Keane' has
                                                    patented several processes for producing  10 to 25 M direct nucleate
                                                    flotation DNF bubbles which should have similar characteristics to
                                                    CGAs, although it is doubtful whether 10 p sized bubbles can have
                                                    more than a transient existence.
                                                      The applications of Colloidal Gas  Aphrons (CGAs) have, to
                                                    date, been focused on the removal of very  fine suspended or
                                                    precipitated panicles. The mechanism of separation can range from
                                                    a combination of ion flotation and precipitation flotation to floe
                                                    flotation, according  to  Sebba and Barnett.4  Barnett  and Liu*
                                                    floated hydrophilic colloidal particles such as found  in fish wastes
                                                    from aqua-culture  systems. Sebba  and Yoon' demonstrated that
                                                    CGAs were able to selectively float coal wetted with kerosene from
                                                    ash.
                                                      Auten and Sebba' have shown  that it is possible to separate
                                                    minerals by selective bubble  entrained  floe-flotation.  They ex-
                                                    plored the ability of fine CGA bubbles to be entrained in the floc-
                                                    culated clay from the phosphate  slime effluent and in this way to
                                                    separate these very finely divided  particles from apatite. In another
                                                    application, CGAs were shown  to be effective in the  flotation-
                                                    harvesting of single ceU algae from a dilute suspension.'
                                                      For many separations, it is the distinctive characteristics of the
                                                    air dispersions which make the  applications technically feasible.
                                                    Sebba and Barnett4 have discussed the use of CGAs for a wide
                                                    range of liquid/liquid and liquid/solid separations.
                                                    In Situ Hazardous Waste Treatment

                                                      The treatment of subsurface  hazardous waste releases to soil
                                                    matrices or to impoundment sediments is limited to: (1) injection/
                                                    recovery techniques with above ground treatment, or (2) subsurface
                                                    in situ treatment (destruction) techniques. This assumes, of course,
                                                    that the releases  exceed the depth appropriate for land  treatment
                                                    and that excavation followed by biodegradation,  destruction, en-
                                                    capsulation and/or disposal in a hazardous landfill is not practical,
                                                    economical or safe.
                                                      In the field, the use of injection/recovery techniques has become
                                                    accepted practice.  For  example, according to a USEPA survey
                                                    completed by Neeley et a/.' of 180 remedial actions (on  169 sites),
                                                    recovery wells and  French drains (with and without surface treat-
                                                    ment) were used  in  15  situations  to  prevent  groundwater con-
                                                    tamination from landfills. On  the other hand, only  two instances
                                                    were noted in which in situ treatment was used, and in each of these
                                                    applications,  the actual groundwater  treatment was carried out
                                                    above ground.
398
ALTERNATIVE TECHNOLOGY

-------
  In a recent update of the remedial action survey, USEPA10 has
reported  that O.H. Materials and the  State of New Jersey  con-
ducted spray irrigation with draw down wells to remove  con-
taminants from groundwater at the Goose Creek site in Plumstead
Township. The actual water  treatment,  again,  was completed
above ground.
  The use of stripping techniques has become quite common for
removing light chlorinated compounds such as TCE and PCE from
contaminated groundwater. For example, Gross and Termaath"
recently reported a study  for the U.S. Air Force at Wurtsmith Air
Force Base treating 6600  mVday. The process included the use of
activated carbon processing for the final treatment following pack-
ed tower aeration. Cummins12 of the USEPA has been conducting
and monitoring tests at several sites to determine the effectiveness
of air stripping for removing contaminants.
  A  survey of the contaminants found at the 114 top priority
Superfund sites has been  completed.13 At these sites, 64 incidences
were noted where slightly  soluble organics, including aromatics and
halogenated  hydrocarbons occurred.  Most  of these would  be
amenable to injection/recovery techniques,  but some  organics
would be biodegradable  if subsurface  in  situ biooxidation tech-
niques were developed and accepted as  "state-of-the-art."
  Occasionally, surface  spills  of  hazardous  waste have  been
treated in place. Harsh14  documented the in situ neutralization of
acrylonitrile by first raising the pH of the area above 10 with lime
and then spraying sodium hypochlorite over  the area. Winn and
Schulte15 reported the cleanup of 6400  kg  of phenol by treating a
total of 3800 m3 of diluted waste with hydrogen peroxide.  Other re-
cent papers by Sikes, et al.16 and Miller and Paddock17 document in
situ treatment of formaldehyde  and acetic anhydride spills respec-
tively. Zitrides1' from Polybac Corporation has provided a bibli-
ography of a number of spill situations  at which full-scale applica-
tions of selected   mutant microorganisms,  as  well as mixed
microorganisms from waste treatment plants, have been used for
biodegradation of surface spills. A more recent paper by Kretschek
and Krupka19 updates the methods and work by Polybac to further
develop the state of the art. Again these citations emphasize above
ground treatment.
Enhanced Injection/Recovery Techniques

  A limited number of studies are continuing to enhance the effi-
ciency of injection/recovery techniques using  surfactant  solubiliz-
ing agents. Given a choice, injection/recovery  procedures followed
by  above ground treatment are usually  preferred to below ground
in situ treatment.
  Texas  Research Institute,20'21 conducted several  laboratory, col-
umn and two dimensional modeling tests on the use of surfactants
to enhance gasoline recovery from high permeability sand. These
results indicated that, with a surfactant combination of an anionic
(Richonate-YLA) oil soluble surfactant and  a nonionic (Hyonic
PR-90) surfactant, effective recovery was  achieved. These results
suggested that  enhanced injection/recovery  techniques  may  be
limited to recovery of slightly  soluble organics released to  high
permeability subsoils. The recovery efficiency also dropped off as
testing went from one dimensional plug flow tests to three dimen-
sional large sand box tests. More recently,  Ellis, et a/.22 completed
laboratory work which has demonstrated overall soil cleanup effi-
ciency of 90% under gravity  flow conditions of intermediate
molecular  weight  aliphatic  and aromatic hydrocarbons,
polychlorinated biphenyl mixtures  and chlorinated phenol  mix-
tures. A combination of  2% each Hyonic PE 90  and Adsee 799,
both nonionic surfactants, were used for testing. Further testing is
in progress.

Subsurface In Situ Hazardous Waste Treatment

  Above ground treatment is generally desirable if the contaminant
can be removed from the soil or sediment matrix  by injection/re-
covery techniques, forced venting or evaporation techniques. The
complete excavation of the site or impoundment sediment followed
by thermal, chemical or biological oxidation treatment, encapsula-
tion, or disposal in a hazardous waste site are also possibilities,
although usually expensive. The final available option is to treat the
buried hazardous waste in place by some chemical, biological or
thermal destruction technique or physical isolation.
  The  treatment of gasoline contaminated  groundwater and soil
matrices using biological techniques has received the most atten-
tion. Recent articles by Lee and Ward23 and  Yaniga24 survey the
work completed. Recently, the American Petroleum Institute (API)
has also initiated a field study on the use of hydrogen peroxide to
transport oxygen subsurface for gasoline degradation.25 In addi-
tion, the  authors have been  testing the use of Colloidal Gas
Aphrons as an oxygen source for subsurface biodegradation under
API sponsorship at Virginia Tech.
  One of the few documented field efforts to conduct subsurface in
situ biodegradation of hazardous organics is  being completed by
Biocraft Laboratories  at its Waldwick, New Jersey site by its sub-
sidiary, Groundwater Decontamination  Systems  (GDS).  The
technology developed  has  been patented and GDS was formed to
sell the biostimulation process to other firms with similar ground-
water problems. Their processing consists of two delivery trenches
(1.5m X 30m x 3 m), upgradient of a contamination source, with
a subsurface drain trench (1.5m x 3m x 24 m) for primary collec-
tion located down gradient. Their innovation comes in the injection
of air at several points into a 3.5 m thick contaminated layer  of
glacial  till and stratified drift located 1 m underground. According
to an earlier soil summary, three basic soil types occurred in the
area: (1) Merrimac gravelly loam, (2) Papakating silt loam, and (3)
muck.  In  addition to some in  situ (subsurface) biodegradation,
above ground biological degradation of the contaminated ground-
water was also conducted before it was recycled back into the rein-
jection trenches.  Nutrient levels and microorganism populations
were closely controlled. The subsurface horizontal flow rate was
estimated to be 0.12 m/day, and it takes an estimated year and a
half for reinjected water to traverse from delivery to drain trench.
Permeabilities (hydraulic conductivities) have been estimated to be
9.4 x  10-9 to 1.7 x 10-5 m/s from slug tests.26
  Finally, laboratory studies have been completed on the use  of
radio frequency in situ heating for decontamination  of hazardous
waste  substance27 and  degradation  of high  concentrations  of
diazinon in soil by parathion hydrolase.28 Whether either of these
treatments  would be  suitable  for deeper subsurface  waste  is
unclear.
  In summary, subsurface in situ biodegradation of organics seems
feasible, but efforts to date have emphasized gasoline degradation.
In addition,  no  effort has been  identified  to enhance in situ
biodegradation of impoundment sediments.
RESULTS AND DISCUSSION
  Earlier small scale laboratory studies have demonstrated that col-
loidal gas aphrons,  when  sparged into  various  unconsolidated
saturated soil matrices, are "picked up"  and retained for a pro-
longed period of time.29 During these tests, 70 to 82% of the air in-
corporated into a CGA made using a nonionic detergent (Tergitol
15-S-12)  immediately adhered to  a saturated coarse  sand after
sparging with a fork-like probe. Thirty days later, 70 to 80% of the
initially  retained CGAs  were still  retained in the saturated sand
matrix as an air dispersion or as coalesced aphrons. On injection,
some expansion of the saturated sand bed occurred and the range
of air to water, by volume, in the saturated sand was 0.56 to 1.25.
With CGAs made using sodium dodecyl benzene sulfonate, a max-
imum of 63% of the CGAs adhered to the coarse sand with about
60 to 70% of these retained after 30 days. On the other hand, direct
air injected following the same procedures rapidly "bubbled up"
and burst through to the surface with little air retention.
  In situ laboratory biodegradation studies also demonstrated that
a combination of CGAs and Pseudomonas putida plus microbial
nutrients could be injected into a saturated anaerobic sand matrix
containing 300 mg/1 phenol solution, and 60% of the phenol was
degraded in 24 hr."
                                                                                    ALTERNATIVE TECHNOLOGY
                                                         399

-------
                          Figure I
           CCA Injection: Pilot Trough TcM (Underway)
                                                                                 Figure 3
                                                                         Injector Delivery Hardware
                          Figure 2
           CGA Injection: Pilot Trough Test (Midway)
                                                                                 Figure 4
                                                                    Injector Details of Plow and CGA Holes
Air Retention in Pilot Trough Tests

  Because of these encouraging laboratory retention studies, scale-
up adhesion tests  were conducted in a 13 cm wide trough, 91 cm
long by 46 cm high with one side made of clear polycarbonate (Figs.
1 and 2). The side view panel made it possible to follow injection
when  plowing  adjacent  to  the near  wall.  The air  colloidal gas
aphrons were generated and introduced below the saturated sand
surface (impoundment bottom). The injector was made from 0.6
cm steel pipe with two 0.3 cm diameter holes located just upwards
from the plow point (Figs. 3 and 4). Injection was made by bracing
the  side supports  to maintain the delivery plow vertical. The per-
cent CGA (gas) retention  could  be determined by knowing the
amount of surfactant solution injected as CGAs, the quality of the
CGAs (percent composition air, determined by withdrawing 100 ml
into a graduate cylinder and allowing slow coalescence to a liquid)
and the water surface level change in the  trough after  CGA injec-
tion. Percent "Air Retained" is  the volume air retained per the
volume air injected (as CGAs).
  The results of these injection studies into Ottawa Federal Fine
silica sand are shown in Table 1. All runs were made with injection
of CGAs about 10 cm below the sand surface, such that the plow
tip was about 13 cm deep (Fig. 3). The "Surfactant" column lists
the  surfactant, concentration and CGA quality. The plow position
was moved to inject at three positions along the 13 cm width.
  The "Total Air Input" was determined by measuring the surfac-
tant liquid solution delivered and CGA quality. The volume "Air
                                                       Retained" was determined by the measured total volume change
                                                       less the total liquid volume added as CGA dispersion. The "%
                                                       Void Volume Filled" was approximated based  on a previously
                                                       measured void volume, porosity of 0.40 for the saturated Federal
                                                       Fine sand and ignoring any expansion of the bed during CGA addi-
                                                       tion. Only the top 10 cm of the saturated sand were considered in
                                                       these calculations.
                                                         These results show that the faster plow rate and the slower CGA
                                                       delivery (flow) rate both contribute to enhanced percent air reten-
                                                       tion. The maximum retention of about 50% may seem low, but the
                                                       crude injector design did not deliver a Tine stream of CGAs through
                                                       the 0.30 cm holes, and this left considerable opportunity for CGAs
                                                       to accumulate and channel up the back of the vertical plow shaft.
                                                       With similar surfactant concentrations, CGA retentions of well
                                                       above 50%  were achieved in laboratory tests in  a  two-inch deep
                                                       sand bed  using  a  pronged  "fork" injector  made  of  0.16 cm
                                                       diameter tubing.  In  addition, 56% void volumes filled with air was
                                                       achieved in  these laboratory tests, again assuming no bed expul-
                                                       sion. The  three-prong  fork (rake) moved through  the sand with
                                                       minimal sand disruption as large gaps, thus assuring better CGA
                                                       contacting with the  sand and thus better retention.
                                                          In summary, large quantities of air (or oxygen) as CGAs can be
                                                       injected and retained in an  unconsolidated saturated sand matrix
                                                       (impoundment)  under  pilot trough  testing. However, better
                                                       plow/injector design and fabrication should be undertaken at the
                                                       field scale to maximize the retention achieved at the laboratory
                                                       level.
400
ALTERNATIVE TECHNOLOGY

-------
                                                           Table 1
                  Results of Adhesion and Retention of Colloidal Gas Aphrons in Saturated Sand Matrix - Sand Trough Tests
Run
No.
1




2


3

4

CGA Flow Rate
(Hater Plus
Surfactant Bubbles)
Tergitol 750 ml/min
15-S-12
0.3 g/1
(59% CGA)
Avg
NaDBS 750 nl/min
(0.3 g/1
(57% CGA)
NaDBS 750 ml/min
0.3 g/1
(58% CGA)
NaDBS 750 ml/min
(60% CGA)
Plow Rate
Medium
30 sec/pass



Med i urn
30 sec/pass

Slow
75 sec/pass
Number of Air Retained I Void
Total Passes Total Air Input Vol. % Vol. Filled
30 (10/position)
30
60


10/position)
Fotal


30 (10/pos.)
30
t>U
10/pos.)
Fotal
20 (10/wall pos.)
10 (center)
30 Total
Slow
75 sec/pass
12 (4/pos.)
12 (4/pos.)
24 Total
5


6

NaDBS 500 ml/min
(592 CGA)

NaDBS 500 ml/min
(59% CGA)
Medium
31 sec/pass

Slow
75 sec/pass
30 (10/pos.)
30
bU
10/pos.)
otal
20 (10/wall)
10 (center)
30 Total
7


8


9

10

NaDBS 500 ml/min
(57% CGA)

NaDBS 250 ml/min
(635 CGA)

NaDBS 250 ml/min
(62% CGA)
NaDBS 250 ml/min
(57% CGA)
Slow
75 sec/pass

Medium
31 sec/pass

Medium
30 sec/pass
12 (4/pos.)
12 (
24 1
4/pos.)
otal
30 (10/pos.)
30
60
10/pos.)
Fotal
30 (10/pos.)
30 (lp/ DOS.)
60 Total
Slow
75 sec/pass
30 (10/pos.)
30 (10/pos.)
60 Total
1758 ml
1895 ml
3653 ml


1640 ml
1694 ml
3334 ml
4350 ml
2956 ml
7306 ml
2847 ml
3072 ml
5919 ml
1701 ml
1436 ml
2737 ml
2888 ml
1168 ml
4056 ml
1396 ml
1644 ml
3040 ml
1581 ml
804 ml
2385 ml
1013 ml
969 ml
1952 ml
2246 ml
2280 ml
4525 ml
660 ml
423 ml
1083 ml


554 ml
46a ml
1022 ml
1015 ml
175 ml
1 1 90 ml
329 ml
417 ml
746 ml
723 ml
433 ml
1T5E
782 ml
293 ml
975 ml
667 ml
590 ml
1257 ml
941 ml
238 ml
1179 ml
529 ml
426 ml
955 ml
1031 ml
480 ml
1511 ml
38%
22%
30T


34%
28%
3T%
23%
6%
T6T
12%
14%
T3%~
43%
29%
47%"
27%
25%
25%"
48%
36%
4T%
59%
30%
49%
52%
44%
W
46%
21%
33%
14%
24%
2~4~%~


12%
22%
22%
22%
26%
26%
7%
16%
T5T
16%
25%
2~5lT
17%
21%
2T%
15%
27%
27%
20%'
26%
?6%
12%
21%
22%
33%
13%

(total)




(total)


(total)

(total)


(total)


(total)


(total)


(total)


(total)

(total)
Aug.
Air Input
Per Pass
59
63



55
56

218
296
237
256

57
48

144
117

116
137

53
27

34
32
75
76
ml
ml



ml
ml

ml
ml
ml
ml

ml
ml

ml
ml

ml
ml

ml
ml

ml
ml
ml
ml
Laboratory Phenol Degradation Studies with CGAs
  To verify early reported results and  establish analytical pro-
cedures, a phenol degradation study in a saturated sand matrix us-
ing CGAs was completed. To a series of 250 ml beakers, 310 g of
Ottawa Federal Fine sand were added. Then 300 mg/1 phenol solu-
tion was added until just covering the sand bed (one pore volume).
Then a 0 .3 g/1 solution of sodium dodecyl benzene sulfonate was
blended into a 33% CGA (poor  quality)  and 78 ml  of liquid
pumped in as CGA. This resulted in an average of 18 cm3 of air as
CGA retained in the saturated sand matrix, approximately 23% of
the pore volume. The CGAs were injected using a 1.0 ml pipe with
vigorous  stirring.  The  phenol concentration in  12  individual
samples was followed and phenol degradation as a function of time
is shown in Table 2. The sacrificed sample was well mixed before
analysis.
  The results verified that after approximately 24 hr degradation
was complete, which is  in agreement with earlier tests." A similar
set of runs was made using 60% quality CGAs, which showed that
95% of the phenol was degraded by 45 hours. Thus, a test period of
24 hr was considered as appropriate for further testing.

Phenol Degradation-CGAs Dispersion/Diffusion

  Both laboratory and  trough tests have demonstrated the ability
of air (or oxygen) as CGAs to be injected in sizeable volumes into
saturated matrices. However, there was concern as to how much
oxygen dispersion occurred during injection as well as how much
diffusion took place subsequently. Since CGAs are injected as a 60
to 70% air dispersion  in water or waste water, some  flow and
dispersion will take  place from the point of injection.  The CGA
dispersion is  also assisted  by the disruption and settling of the
saturated matrix  or impoundment sediment during injection.
  Small scale pilot dispersion studies were conducted in a 28 cm
diameter, 10 cm  high glass vessel. Federal sand was placed in the
vessel (5 cm deep.) and the sand saturated to the sand surface with a
                           Table 2
    Degradation of a 300 mg/l Phenol Solution in a Saturated Sand
                  Matrix as a Function of Time
                    (Laboratory Batch Test)
   Elapsed Time
       (hr)

        0
       4.8
       8.2
       20.5
       28.1
       47.3
Degradation
   (%)
 0(±10%)
    0
    0
   80%
   100%
   100%
No. Samples
 Sacrificed

     2
     2
     2
     2
     2
     2
300 ppm phenol solution.  CGAs containing Pseudomonas putida
and nutrients were introduced through a 1 ml pipette injector into
an 8 cm diameter center section of sand. All CGAs were made using
a 0.3 g/1 NaDBS solution to produce a 50% quality CGA disper-
sion.  Caution  was taken  to  keep the  initial phenol  solution
anaerobic, and following CGA introduction the top of the vessel
was swept with carbon dioxide and covered to eliminate oxygen dif-
fusion to the water surface.
  The results of these dispersion/diffusion studies are shown in
Table 3.  The initial concentration of the phenol was adjusted to
reflect dilution by water introduced in the CGAs added to the 300
mg/1 phenol charge solution. The final phenol solution was deter-
mined by mixing and draining the treated product from the test
vessel and measuring the phenol using a colorimetric assay, based
on rapid condensation of phenol with 4-aminoantipyrene, followed
by oxidation with  potassium ferricyanide. The gas hold-up was
estimated using relatively  inaccurate volume measurements, and
thus the gas hold-up could be in error by ± 50 ml.
                                                                                   ALTERNATIVE TECHNOLOGY
                                                         401

-------
                          Table 3
    Phenol Degradation Using Injected CGA as Oxygen Source
                  (24 Hour Diffusion Study)
» CGA
Gu



A Air
B Air
C Air
D Air
E 02
F 02
E>t. Gu
Hold-Dp
(ml)


98
198
205
121
173
80
Inlllil Vol. Llq.
Added (ml)
|«>Gu|


1710 (6)
1670 (12)
1530 (131
1500 (8]
1470 [ 121
1480(51
Inlllil Flnil
Phenol Phenol
Cone. Cone.
(Dilution mg/1
Adjuil.)
280 220
265 240
270 210
280 200
260 200
290 230
Blode-
grtdt-
(Ion
(V»)

21
9
22
29
23
21
   If the CGAs were effective only in the centered 8 cm diameter in-
 jection region of the 28 cm diameter vessel, the maximum phenol
 degradation would have been 7.4%. The results given in Table 3 in-
 dicate the effectiveness was 3 to 4 times greater. In addition, the
 estimated CGA hold-up gas in the  injection region was equivalent
 to 5% to 13% of the total pore volume of the vessel, ignoring any
 expansion. Thus,  in  these tests,  anywhere from about 1 to 4
 volumes of saturated water containing phenol were treated for each
 volume of CGA air or oxygen retained in  the saturated sand.
 Phenol Degradation-Trough Test

   The final test in this sequence was a phenol degradation study in
 the test trough. Specific concern centered on delivery of CGAs and
 biodegradation of the 300 mg/1 phenol solution. The details of the
 test are shown in Table 4.

                            Table 4
            Biodegradation of Phenol Using Air CGAs Plus
                   Microorganism and Nutrients
                        Pilot Trough Test
                                        12380 ml (300 mg/1 phenol)
                                      = 1320 ml (30 mg/1 phenol)

                                      = 2242 ml
                                          14792ml
                                          ) (240 mg/1 phenol)

                                      = 2550ml
1. Total initial phenol solution to saturate
  15 cm federal fine sand and provide
  5 cm liquid head
2. Total liquid added as CGA (100 ml of
  a P. putida seed culture was added
  to 5.01 of 0.3 g NaDBS/1  solution
  for CGA generation)
3a. Liquid removed so as to return liquid
   head to original 5 cm level
3b. Liquid removed so as to reduce
   liquid heat to 0.6 to 0.9 cm
4a. After 24 hr, five 10 ml samples with-
   drawn at different positions along the
   box, and all 2.5-k3.8 cm deep, and
   analyzed for phenol concentration
   (top row of concentrations shown
   below)
4b. Five 10 ml samples withdrawn, as
   above, 5 to 64 cm deep (lower row of
   concentrations)
5a. Upper sand bed collected, mixed, and
   analyzed
5b. Lower sand bed collected, mixed, and
   analyzed
          Test Trough Schematic and Point Sample Analysis Result!

   Liquid Hnd
6 in
4 in
2 in
_ Sand
_ 190

~ 250
— —Injector port level
— —Plow lip level

140 210 190 250
(mg/1)
180 200 190 260 -1
(mj/l)_
—
T
Upper
Sand
Bed

1
Lower
Sand Bed
   The COAs were added at a flow of 250 ml of CGAs per minute
 with a slow plow rate of 75 sec per plow pass. A total of 20 passes
 were made at the  10 cm injection depth—10 at the middle  position
 and 5 apiece at each side (wall) position. With a 61% quality COA
 solution, 1320  x  61/39 or 2065 ml of air was injected and 922 ml
 retained, for a 45% retention. Given that air was distributed into
 the top 11 cm, then 922 ml of a total pore volume of approximately
 8,000 ml was occupied (12%).
   The in situ biodegradation performance is also of interest. After
 point concentration samples were taken (Table 4), the upper sand
 bed (plus liquid)  was  skimmed, thoroughly mixed and the water
 analyzed for phenol yielding a 230 mg/1 average concentration. The
 922 ml  of  retained  air provided about  254  mg/l  of O2 for
 biodegradation  @ 25 °C. Furthermore,  assuming that only the
 phenol in the upper 11 cm of sand was biodegraded, then a total of
 approximately 6.7 1 of water were treated with a 292 to 230 mg/1
 reduction in phenol concentration. The total phenol biodegraded
 was 414 mg of phenol. Thus, 1.6 mg phenol/mg O2 retained were
 degraded in  24  hr. The degradation agrees with  typical yields of
 substrate consumed per O2 utilized, which typically ranges from 0.5
 to 3.0."
 CONCLUSIONS
   Given that hazardous treatment is to be conducted  in situ, the
 use of CGAs offers a unique opportunity to introduce oxygen, or
 perhaps  ozone, to treat dissolved and possibly suspended organics
 found  in saturated impoundment sediments. Delivery of CGAs by
 some injection  plow looks especially promising for converting an
 anaerobic impoundment to an aerobic environment for enhanced
 biodegradation. Under such  circumstances, air injection is totally
 ineffective.
   Delivery of CGAs by a series of injection wells could also prove
 effective for introducing air or oxygen into saturated subsoils, into
 partially saturated  subsoils  at the water interface where  floating
 organics will concentrate and perhaps into the capillary  zone above
 the saturated region. The adhesion and  retention of  CGAs and
capability to mix and deliver  a complete treatment solution (air or
oxygen, as CGAs,  nutrients and selective microorganisms) to a par-
ticular area provides a one-step  biodegradation  formulation.
 ACKNOWLEDGEMENT

  The  authors appreciate the partial support of the U.S. AFOSR
and particularly the Environics Division of the AFESC located in
Tyndall AFB, Florida in carrying out this project.  We are also
greatly  indebted to T.F. Douglas and R.S.  Paris  and recent
students   at  Virginia  Tech  for  assistance  in  conducting the
laboratory and  pilot testing.

 REFERENCES

 1. Auten, W.L.  and Sebba, F., "The  Use  of Colloidal Gas Aphrons
    (COAs) for  Removal  of Slimes  from Water by  Floe-Flotation,"
    "Solid and Liquid Separation," John Gregory, ed., Ellis Horswood,
    England, 1984, 41.
 2. Sebba, F., "Microfoams.  An Unexploited Colloidal  System,"  J.
    Coll.  A Interface Sc., 35. 1971, 643.
 3.  Sebba, F., Investigation of the Modes of Contaminant Capture in
    COA (MOD) Foams, Report to O.W.R.T., 14-34-0001-0489, October,
    1982.
 4. Sebba,  F. and  Barnett, S.M.,  "Separation Using Colloidal Gas
    Aphrons," Proc, 2nd Int. Congress ofChem. Eng., IV,  1981, 27-31.
 5. Keane, J., "Treatment  of Waters with Broad  Spectrum Contamin-
    ants," U.S. Patent 4,417,985, Nov. 29, 1983.
 6. Barnett, S.M. and Liu, S.F., Proc. Conference on Seafood Waste
    Management in  the 1980's, 1980.
 7. Sebba, F. and Yoon, R.H., "The Use of Micron-Sized Bubbles in
    Mineral Processing" in "Interfacial Phenomena in Mineral Proces-
   sing,  B. Yarar and D.J. Spottiswood,  eds., 1982, 161,
 8. Honeycutt, S.S., Wallis, D.A. and Sebba, F., "A Technique for Har-
    vesting Unicellular Algae Using  Colloidal Gas Aphrons "  Biotech
    and Bioengr. Symp. No. 13, 1983. 563-575.
402
         ALTERNATIVE TECHNOLOGY

-------
 9. Neely, N.D., Schauf, F. and Walsh, J., Remedial Actions at Hazard-
   ous  Waste Sites—Surveys and Case Studies. EPA-430/9-81-05 SW-
   910, 1981.
10. U.S. Environmental Protection Agency, Case Studies 1-23: Remedial
   Response  at Hazardous  Waste Sites, EPA-540-2-84-002b, March,
   1984.
11. Cross, R.L. and Termaath,  S.G.,  "Packed Tower Aeration  Strips
   Trichloroethylene from Groundwater," Presented at A.I.Ch.E. Meet-
   ing, Philadelphia, PA, Aug. 21, 1984.
12. Cummins, M.D., Personal Communication (1982).
13. EPA Environmental News, 1981.
14. Harsh,  K.M., "In  Situ Neutralization of an Acrylonitrile  Spill,"
   Proc. of 1978 National Conference on Control of Hazardous Ma-
   terial Spills, Miami, FL, Apr. 1981, 187-189.
15. Winn, B.M. and Schulte, J.H., "Containment and Clean-Up of a
   Phenol Tank Car Spill, May-June 1978, Charleston, SC," Proc. 1982
   Hazardous Material Spills Conference, Milwaukee, WI, Apr., 1982,
   11-14.
16. Sikes, D.J., McCulloch, N.M. and  Blackburn, J.W., "The Contain-
   ment and Biological In Situ Treatment Techniques, Proc.  1984
   Hazardous Materials Spill Conference, Nashville, TN, Apr., 1984,
   38-44.
17. Miller, J. and Paddock, J., "Acetic Anhydride Spill at Thorp, Wis-
   consin," Proc. 1984 Hazardous Material Spills Conference,  Nash-
   ville, TN, Apr. 1984, 45-49.
18. Zitrides, T.G., "More on Microorganisms." Environ. Sci. Technol.,
    16, 1982, 431A-432A.
19. Kretschek, A. and  Krupka,  M., "Biodegradation as  a Method of
   Hazardous Waste Treatment in Soil and Subsurface Environments,"
   Proc. Conference  on Hazardous  Waste and Environmental  Emer-
   gencies, Houston, TX, Mar., 1984,  220-226.
  ,
/2H
V_y
/20J Texas Research Institute, "Underground Movement of Gasoline on
     Groundwater and Enhanced Recovery by Surfactants," Submitted to
     the American Petroleum Institute, Sept.  1982.
 2H Texas Research Institute, "Test Results of Surfactant Enhanced Gas-
     oline  Recovery in  Large-Scale Model Aquifer,"  Submitted to the
     American Petroleum Institute, Apr., 1982.
 22/ U.S. Department of  Agriculture, Soil Taxonomy: Basic System  of
 —  So/7 Classification for Making and Interpreting Soil Survey, Agri-
     culture Handbook No. 436, 1975.
 23. Lee,   M.D.  and Ward,  C.H., "Reclamation  of  Contaminated
     Aquifers: Biological Techniques," Proc. 1984  Hazardous Material
     Spills  Conference, Nashville, TN, Apr., 1984, 98-103.
 24. Yaniga, P.M., "Groundwater Abatement Techniques for  Removal
     of Refined Hydrocarbons," Proc. Conference on Hazardous Wastes
     and Environmental Emergencies, Houston, TX, Mar., 1984, 374-383.
 25. Carlson, C., Personal Communication, July, 1984.
 26. Jhaveri, V. and  Mazzacca, A.J., "Bio-Reclamation of Ground and
     Groundwater," Proc. National Conference on Management of Un-
     controlled Hazardous Waste Sites, Washington,  DC,  Nov., 1983,
     242-247.
 27. Harsh, D., Bridges,  J.E.  and  Sresty, G.C., "Decontamination of
     Hazardous Waste Substances  from Spills and  Uncontrolled Waste
     Sites by Radio  Frequency In Situ Heating." (incomplete)
 28. Paulson,  D.L., Honeycutt, R., Lebaron, H. and Seim, V., "De-
     gradation of High Concentrations of  Diazinon in  Soil by Parathion
     Hydrolase", Proc.  1984  Hazardous Material Spills  Conference,
     Nashville, TN, Apr.,  1984, 92-97.
 29. Michelsen, D.L., Wallis, D.A. and Sebba,  F.,  "In Situ Biological
     Oxidation of Hazardous Organics," Env. Prog.,  3, 1984, 103-106.
 30. Bailey, J. E.  and  Ollis,  D.F., Biochemical  Engineering Funda-
     mentals, McGraw-Hill, Co., New York, NY, 1977, 479.
                                                                                          ALTERNATIVE TECHNOLOGY
                                                               403

-------
                     THERMAL TREATMENT OF SOLVENT
                                   CONTAMINATED  SOILS

                                                DIANE  HAZAGA
                                                   SUE FIELDS
                                    U.S. Environmental Protection Agency
                                  Emergency and Remedial Response  Branch
                                                 Atlanta, Georgia
                                         GARY P.  CLEMONS, Ph.D.
                                    Hazardous Waste  Technology Services
                                                 Atlanta, Georgia
INTRODUCTION
  Hazardous waste sites frequently contain large quantities of con-
taminated soils. The costs associated with transporting and dispos-
ing these contaminated soils to an approved landfill can be prohib-
itive when large volumes of soil are involved. Additionally, land-
fill may not provide a permanent solution to the problem; the soils
may have to  be handled again at some future date to effectively
render them non-hazardous. Thermal treatment of contaminated
soil can be a viable alternative.  The following material is  the
method description and decision rationale for the treatment of sol-
vent contaminated soils from an abandoned solvent reclamation
facility.

BACKGROUND
  The subject site for this report  is one acre in size and located
in a heavily industrialized area. The facility operated from 1971 to
1983; spills and discharges significantly contaminated local soil
and ground water. As a complicating factor, a drinking water well
field which serves 100,000 people is located approximately 2,100
ft from the site.  During  1984,  the USEPA instituted  a  cleanup
action to mitigate the threat of solvent migration into the well
field.
  Contaminated soil on-site is classified as brown and white sandy
topsoil; groundwater is encountered between 4 to 7 ft  below  the
soil surface throughout the site. The major contaminants found to
be flowing from the topsoil to the groundwater surface were 1,1,1
trichloroethane, trichloroethene, toluene, ethyl benzene and ortho-
meta- and para-xylene in concentrations ranging from 10,000 to
110,000 ppb.

RATIONALE

  Two alternatives for mitigating the contaminated soil problem
were found to be feasible. The first alternative would involve ex-
cavation of the contaminated soils,  transportation to a secure haz-
ardous waste landfill for disposal and backfilling of the excavated
areas.
  The second alternative would entail excavation of contaminated
soil, transportation to a thermal treatment unit  where  the  soil
would be rendered non-hazardous and return of the treated soil to
the excavated area. This alternative would also require some addi-
tional backfilling on the site. The criteria affecting the decision to
thermally treat the soil were based on contaminant  properties  and
concentrations and the cost of transportation  and  disposal to an
approved landfill.  Based on these  criteria, the soil contaminants
had to be easily volatilized, be so diluted after treatment that no
threat to the  public or the environment exists and be in such  low
concentrations that  volatilization would not require  excessive
energy consumption and therefore higher operating costs.

PRETREATMENT PROCEDURES

  To determine if the thermal treatment of site soils would work
satisfactorily, a trial run was conducted using a pilot scale batch
type heater that would accommodate approximately 0.25 yd' of
soil. Two test burns using the site soils were conducted. Chemical
analyses before and after burning for  1,1,1 trichloroethane, tri-
chloroethene and toluene showed greater than 99% removal for
each contaminant. From this trial, it was determined the soil would
have to be heated to 375 °F for a contact time of approximately
2 mm for effective removal. It was also learned that paniculate
emissions would be a problem with a full scale system.
  The next task was to apply for an air discharge permit from the
state. The state of Florida reviewed the unit specifications for air
discharge permit requirements. Paniculate emissions were to be re-
moved by a cyclone followed by a cloth bag house. The air to cloth
ratio was calculated for effective paniculate removal. The emis-
sions collection system was rated at 99% paniculate removal effic-
iency by the manufacturer. Based on the air to cloth ratio and the
99% efficiency rating, the state approved the paniculate collection
system.
  The expected  volatile organic  carbon (VOQ discharge levels
were calculated from  contaminant concentrations, soil feed rate
and discharge gas flow rate. Worst case VOC discharge concen-
trations were assumed. To insure that the process did not pose a
threat to the surrounding environment or the personnel operating
the unit, an air monitoring program would be instituted during full
scale operations.
  The full scale thermal treatment unit met all the applicable air
emissions requirements and the Florida DER  advertised the pro-
posed operating  permit in the local newspaper as required. After
the two  week  public opposition response time had  expired with-
out adverse comments, the  state issued a temporary operating
permit to the USEPA and the unit was mobilized.

TREATMENT UNIT DESCRIPTION

  The contaminated soil was treated in an asphalt drying unit de-
signed to dry raw materials  as they enter an asphalt plant. The
system (Fig. 1) included the asphalt preheater,  a paniculate collec-
tion system and a 500 KW  generator for all electrically powered
components on-site. A schematic block diagram of the treatment
process is provided in Figure 2.  The major components of the sys-
tem are  commercially available and only minor  fabrication work
was required.
404
         ALTERNATIVE TECHNOLOGY

-------
                                                           Figure 1
                                                        Actual Treatment
  The preheater was propane fueled and designed for a feed rate of
100 tons/hr, but was only operated at a rate of 10-15 tons/hr to
comply with air discharge requirements. Liquid propane contained
in a 5,000 gal tanker was sent through six 70 gal/hr vaporizers be-
fore being ignited at the preheater inlet. Unit operating parameters
are given in Table 1.

PROCESS PROCEDURE
  Because of limited space on-site, off-site treatment was required.
Contaminated soil was excavated from  the site and  transported
via 15 yd3 dump trucks to the treatment unit located  at a county
landfill  in  a remote area approximately 15 miles away.  When
necessary, the soil was initially spread in a contained area to dry
and then placed into the collection hopper of the treatment unit.
From the hopper, the soil travelled via a belt-driven conveyer into
the preheater at a rate of 10-15 tons/hr. On the average, two dump
trucks carried 130 tons of soil per day to the unit for treatment.
  The soil was gravity fed through the rotating preheater; the fuel
was burned at the other end. Agitated by the rotating action of the
preheater and exposed to gas at 375 °F, the organics vaporized from
the soil. Since the organic contaminants included combustible com-
pounds, it is reasonable to assume  that combustion as well as vola-
tilization occurred during the treatment process.
  The treated soil was collected in a bin and was characterized as
a dry sand. To improve the  consistency and add natural organic
content, the soil was mixed with fill dirt from the landfill at min-
imal cost. The mixture was then returned to the site. For the entire
tonnage of soil treated (1,670 tons), the average loss in soil weight
was 20%. The majority of this weight reduction was likely due to
moisture evaporation and charring  of natural soil organics.
  Participate emissions from the treatment unit were  collected in
the cyclone  and the cloth bag house. Some volatile  material emis-
sions were discharged into the surrounding  atmosphere. Air mon-
itoring was  conducted during the second day of regular treatment
operations to determine stack emissions and ambient air concentra-
tions. Air samples to determine VOCs were collected from the stack
gas, one position 900 ft upwind, and nine positions  approximately
900 ft downwind in the radius of the stack plume path.
RESULTS
  The treatment unit was used to effectively treat 1,670 tons of
contaminated soil by reducing the concentration of volatile organ-
ics  present. Concentrations of  1,1,1 trichloroethane,  trichloro-
ethene, toluene and xylene were reduced by at least 99% (Table 2).
  Since the treated soil was returned to the site as fill, both  dis-
posal and major backfilling costs were avoided. Transportation of
the untreated soil to the treatment site was included in the treat-
ment cost, while the return of treated soil was excluded. The treat-
ment price did not include operators and equipment for soil exca-
vation and backfilling. Soil loading and unloading operations at the
treatment site were also excluded.
  The low level concentrations of organic emissions discharged to
the atmosphere posed no threat to public health or  the environ-
ment (Table 3).  Stack samples contained VOC concentrations of
5,000 to 20,000 ppb for the major contaminants. The downwind
samplers which intercepted the plume at ground level samples  had
a maximum level of only 8 ppb.
  Effective volatilization of the contaminants required 6-10 gal of
propane fuel per ton of soil. Thirteen operating days were needed
to complete the soil treatment.
                BAG HOUSE  AM)

                MR DECHARGE
SOL muFTTOH
an


PRBCATJER WITH
CYCLONE CHAMBER


FEED HOPPER
TO CONVEYOR
                           Figure 2
                 Flow Chart of Treatment Process
                                                                                    ALTERNATIVE TECHNOLOGY
                                                          405

-------
                            Table 1
               Treatment Unit Operating Parameters
   Parameter

   Feed rate
   Operating Temperature
   Total heat input
   Stack gas discharge rate
                       (^aerating Conditions

                            10-15 tonsAir
                              375 «F
                            18.5 M tmi/hr
                         22,725 ft3 /mm, 77  ft/six-
                            Table 2
   Concentration of Major Contaminants In Untreated and Treated Soil
Chemical                    Untreated 
-------
       FLOATING COVER  SYSTEMS FOR  WASTE LAGOONS:
POTENTIAL APPLICATION AT OLD  INGER SITE, LOUISIANA

                                                MARK L. EVANS
                                                JOHN P. MEADE
                                                  JRB  Associates
                                                McLean, Virginia
                                            ANTHONY N . TAFURI
                                     U.S. Environmental Protection Agency
                                  Oil and Hazardous Materials Spills Branch
                                                Edison, New Jersey
 INTRODUCTION

   There are many pits, ponds and lagoons at uncontrolled haz-
 ardous waste sites across the nation. With sufficient rain, these im-
 poundments can overflow, resulting in the release of their chem-
 ical-laden contents to the environment. One approach to this prob-
 lem involves pumping, transporting and treating enough of the im-
 pounded material to create a safe margin of freeboard between the
 remaining liquid and the top of the containment structure. These
 pumpdown/treatment measures are considered temporary, how-
 ever, and are usually initiated to prevent damage to the environ-
 ment  while the site is undergoing investigation and ranking for
 eventual remedial action.  They also can be expensive, costing be-
 tween $0.10 and $0.40/gal', and are especially expensive if pump-
 down/treatment is required several times before a final action can
 be undertaken at a site.
   To address this problem, the Oil and Hazardous Materials Spills
 Branch (OHMSB) of the USEPA assembled cost and technical in-
 formation on methods to prevent rainwater from entering these
 impoundments. The resulting report2 on spillage  control methods
 for waste  lagoons concluded that conventional structural covers
 (i.e., metal buildings, air supported structures and cable  struc-
 tures) are  not generally cost-effective for preventing overtopping
 of waste lagoons unless there are unusual  circumstances increas-
 ing pumpdown/treatment costs. The report recommended the in-
 vestigation of innovative structures, such as floating covers, as an
 alternative to pumping, transport and treatment of impounded
 liquids.
   In response to the above-mentioned recommendation  and a re-
 quest from USEPA Region VI for information on this technology,
 OHMSB directed JRB  Associates to: (1) perform a state-of-the-
 art review of floating cover systems and (2) recommend specific
 alternatives to prevent the potential overtopping of a waste lagoon
 at the Old Inger site near Darrow, Louisiana. This paper docu-
 ments the result of these efforts and includes information obtained
 from actual site visits and from 15 cover/liner manufacturers and
 installers3'20.
 TECHNOLOGY DEVELOPMENT AND CURRENT USES

   Floating  covers were  initially  developed as an  inexpensive
 method for covering potable water reservoirs to eliminate contam-
 ination  from surface runoff and organic debris. The first patent
 was issued in 1967 to Globe Linings, Inc. for a design that allowed
 extreme fluctuations in liquid levels and drainage of accumulated
 rainwater from the cover surface.28
  These covers were installed in many reservoirs during the late
1960s and early 1970s by various installation companies across the
country. The first covers were made of either butyl rubber or a syn-
thetic polymer called Ethylene-propylene-diene-monomer (EPDM)
and were fitted with butyl rubber floats that required inflation.12-26
Current floating covers are composed of Hypalon or chlorinated
polyethylene (CPE); however, other materials are also used for
special projects.3'w The old air-inflated rubber floats have been re-
placed by closed cell polyethylene floats.
  In 1976, two additional patents were  issued: one was issued to
Burke  Industries, Inc.; the other  was  issued to Globe Linings,
Inc.29'30 The Bruke patent presented a  new method for draining
rainwater from a floating cover, while the Globe patent presented a
new system for collecting gases from beneath covers placed over
biodegradation ponds.
  At this time,  Globe  Linings,  Inc. has installed approximately
200 covers, while Burke and five other companies have installed
about 100 more.12'16 These covers have  ranged in size from about
15,000 to about 700,000 ft2; however, larger covers also have been
installed. One cover, now under construction in the Los Angeles,
California area, covers 2,000,000 ft2.3-20
  According to industry officials, approximately 80 to 85% of all
floating covers  have been constructed over potable water res-
ervoirs.12'16 The remaining 15  to 20% have  been used on biode-
gradation facilities, slaughterhouse waste lagoons, chemical treat-
ment ponds, toxic waste lagoons and fish hatchery ponds.12'16 The
covers for biodegradation facilities were  developed to collect meth-
ane gas for use in boilers or electric generators.  Odor control was
cited as the main reason for covering pits at slaughterhouses. Fish
hatchery ponds were covered to prevent fishkilling algal blooms.
Chemical treatment ponds and toxic waste lagoons were covered to
prevent overtopping due to precipitation. Due to the proprietary
nature of these activities, very little data were available on the types
of chemicals or facilities involved in covering waste lagoons.
  Regardless of where they have been placed, floating cover sys-
tems have shown a remarkable resistance to the elements at a very
low cost. Average costs ranged  between $2.50  and  $3.40/ft2;3-20
much lower than other available methods for covering large bodies
of liquid.2 In addition, many of the representatives stated that a 20
year guarantee was provided for their system.6'1-12'19'20 The dur-
ability  of floating covers is exemplified by a cover installed on a
potable reservoir in Fort Saint  John,  British Columbia12 which
withstood the forces of 4 ft of ice and  -40°F temperatures for 2
months. (The temperature dropped to -65°F for 3 weeks during
that 2 month period.) Two other covers in Mobile, Alabama have
been unaffected by winds up to 150 miles/hr.'8
                                                                              ALTERNATIVE TECHNOLOGY
                                                      407

-------
DESIGN OF FLOATING COVERS

  As mentioned in the previous section, there are two patented de-
signs for floating covers and an additional patent for the associated
gas  collection system.  These designs, with slight modifications,
have been utilized in virtually every floating cover system installed
to date.

The Globe Design
   The Globe design is primarily based on a rainwater collection
sump that forms around the perimeter of the cover during a rain-
fall. Rainwater is directed toward this sump by various configura-
tions of foam floats inserted and sealed into special pockets in the
cover. The most common configuration consists of one long center
float attached perpendicularly to smaller, lateral floats at about 36-
ft centers (Fig. 1). After the perimeter sump fills with water, it is
drained either by random pumping with a portable pump or by in-
stalling a flexible perforated hose in the area of the fold formed by
the  collected water. On large reservoirs, vertical cables have been
strung  from the reservoir bottom and attached to the cover to
stabilize it against wind action.
   The patented Globe gas collection system is designed for connec-
tion to the foam floats on the Globe cover. However, since all cover
designs use the same type of floats, the Globe gas collection system
could be fitted to almost any floating cover. The gas collection sys-
tem consists of several air chambers formed at various points along
the cover flotation system (Fig. 2). Each chamber is covered by the
 floating cover  material and is equipped with openings that allow
gases to flow  up  through the air chamber  from the triangular-
shaped space formed between the cover, the covered liquid and the
side of the float. The upper end of the air chamber is connected to
a short standpipe and then to a manifold pipe to allow pumping of
gases.
                          Figure 1
Schematic Overview and Cross-Section of a Floating Cover Incorporating
                   the Patented Globe Design
                                                                                  Figure 2
                                                          Schematic Overview and Cross-Sections of a Patented Globe Floating
                                                         Cover and Gas Collection System Design (Redrawn from Kays," Patent
                                                                              Number 3,980,199)

                                                        The Burke Design
                                                          The Burke design,  presented  in  November, 1976, presents a
                                                        completely different method of draining rainwater from the float-
                                                        ing cover. It consists of channels in the middle of the cover created
                                                        by strings of segmented sand-filled tubes that are held at a constant
                                                        depth beneath the cover by floats on either side of the channel (Fig.
                                                        3). The sand-filled  tubes lie underneath and are connected to flex-
                                                        ible perforated collection tubes. Rainwater, after striking the cover,
                                                        drains through openings in the floats and into the collection chan-
                                                        nels; then it is pumped off through the collection tubes. The dimen-
                                                        sions for each component of the system are determined through
                                                        a few simple equations that are based on the dimensions of the area
                                                        to be covered.  With the exception  of the  sand-filled tubes, the
                                                        Burke design can be built with the same materials used  in the Globe
                                                        design.

                                                        MATERIALS OF CONSTRUCTION
                                                          About 90% of all floating covers are made with reinforced Hy-
                                                        palon, either 36 or 45 mils in thickness.3-20 Hypalon is used because
                                                        it is  resistant to most chemicals  and weather conditions, easy to
                                                        seam in the field and among the least expensive of the potential
                                                        material. After  Hypalon was developed, EPDM and butyl rubber
                                                        were  abandoned due to cracking  and  seaming  problems.  The
                                                        second most popular material for floating covers is reinforced 36 or
                                                        45-mil CPE." This material  costs  about the  same  as Hypalon
                                                        ($0.50 to $0.70/ft2) but is neither as easy to seam in the field, nor
                                                        as durable. Because of these drawbacks, the use of CPE covers is
                                                        declining."
                                                          The main limitation of Hypalon, i.e., its low resistance to oil de-
                                                        rivatives such as kerosene, diesel fuel and hydraulic fluids, led to
                                                        the development of three other membranes.  Two of these, used by
                                                        Globe Linings,  Inc.,  are  called OR-EPA and  HR-EPA." Globe
                                                        had  no organized  test data on either of these materials and de-
                                                        clined to release the name of the manufacturer; however, a sample
                                                        of OR-EPA appeared to be more resistant to stretching than sam-
                                                        ples  of Hypalon. A representative of Globe stated that HR-EPA
                                                        is  stronger and resists more  chemicals  than OR-EPA,  which is
408
ALTERNATIVE TECHNOLOGY

-------
 Point of
  Cow
Atttchnwit
                                        Flexibto Collection HOM
                                        Ssnd-FilKd W«ighl
                          Figure 3
      Schematic Overview and Cross-Section of a Floating Cover
  Incorporating the Patented Burke Design (Redrawn from Burke, et al.
                 1976, Patent Number 3,991,900)
stronger and resists more chemicals than Hypalon.12 Those tougher
materials  are  priced accordingly, with HR-EPA  costing about
$1.70/ft2 and OR-EPA costing about $1.50/ft2.12 The Globe rep-
resentative also stated that HR-EPA and OR-EPA are more diffi-
cult to field seam than Hypalon.'2
  The other alternative to Hypalon is XR5 made by the Shelterite
Company. This material is highly resistant to a wide range of chem-
icals but, like the other alternatives to Hypalon, XR5 is more diffi-
cult to field seam and is about twice as expensive.

APPLICABILITY OF FLOATING COVERS AT
THE OLD INGER SITE
  The Old Inger Site is located near Darrow, Louisiana, between
Louisiana Highway 75 and  the Mississippi  River levee  (Fig. 4).
Once an oil reclamation plant, this site was abandoned in 1978
and is now on the National Priority List. It has been the subject of
a remedial investigation and will receive a state-led remedial action
in late 1985 or early 1986.2'25
  The site contains a waste lagoon (Fig. 4),  approximately 240 ft
long and 150 ft wide, which has received several wastes from oil re-
fining processes. This lagoon is only accessible  on foot except for
the northeast corner where a light truck may be able to reach the
edge of the berm. Presently, the lagoon is filled to within 6 in. of
the top; the liquid has three layers:
•A top layer, approximately 1.5 ft thick, composed of a floating oil
 emulsion
•A middle layer of oil-contaminated  water that is from  3 to 8 ft
 thick
•A bottom layer of unknown  thickness, composed of oil sludges
 and debris
  The top layer has a thin hardened film that allows small objects
such as rocks and sticks to  roll across the lagoon surface when
thrown. This thin film apparently allows rainwater to penetrate and
gases to escape, but it does not  permit any appreciable evaporation
of water from the lagoon. Piercing through this top layer are long
2 x 4 in. beams, several  hundred small bottles and approximately
50 55-gal drums. An analysis of chemical compounds contained in
the lagoon revealed the presence of a large number of petroleum-
derived  compounds  including a  series  of long-chain saturated
hydrocarbons,  polynuclear  aromatics, phenols, aromatics  and
phthalate esters.32
                                                                                                         Large storage
                                                                                                         tanks
                                                                 ,-LIVII B0»0
                                                          Figure 4
                         Old Inger Refinery, Site Location and Site Plan (From Law Engineering Testing Company)
                                                                                   ALTERNATIVE TECHNOLOGY
                                                                                                                         409

-------
  The berm around the lagoon is approximately 5 ft high and 8 to
12 ft wide at the base. The width at the top is between 1 and 3 ft,
just barely wide enough to allow one person to walk the perimeter.
Materials used to construct the berm appear to consist of local
earth (a silty, sandy clay).  The average vertical and horizontal
permeability of these soils is about 1 x 1Q-5 cm/sec (10 ft/yr)."
It is not known whether any compaction techniques were used dur-
ing construction or if any efforts were made to prevent sand seams
from forming.
  Furthermore, the north and west outside toe of the berm is cov-
ered with approximately 1 to 1.5  ft of water from  an adjacent
swamp. This swamp, which was contaminated  by a breach in the
lagoon berm in 1978, extends about 1500 ft north of the lagoon
and  connects with a large pasture  by an  overflow  pipe that  runs
under Louisiana Highway 75. Since the 1978 spill, the swamp has
periodically been contaminated by lagoon fluids seeping through a
deep notch  in the top of the berm.21 This notch has recently been
repaired.
  Until November, 1983, Federal and state officials believed that
the liquid level in the lagoon was fairly constant, subject only to
overtopping from rainwater." In July of 1983, an emergency ac-
tion was conducted at the site to create about 2 ft of freeboard by
pumping approximately  300,000 gal of the lagoon's liquid into
three, on-site tanks. However, in late November, the area received
several inches of rain and the fluid level of the lagoon dropped
about 3 in." Presently, there is no confirmed explanation for  this
drop which is still under investigation. Prior to the drop in fluid
level, studies were being made on alternative approached to prevent
the lagoon at Old Inger from overtopping.

Cost of a Floating Cover at Old Inger

  Initial cost estimates for installing a floating  cover over the Old
Inger lagoon were low, i.e., between $2.50 and  SS.SO/ft1.3'20 How-
ever, when  all the site conditions were evaluated and provided to
11 cover installation companies, only one cost  estimate  was re-
ceived;12 it was for $5.50 to $6.50/ft2. Assuming a design area of
approximately 36,000 ft2  for the Old Inger site, the cost of a float-
ing cover would be approximately $250,000. This includes the OR-
EPA material cost (about $1.50/ft2); the combined  costs of the
patented float system and gas collection system (about  $l/ft2);
the installation costs for field seaming, float insertion and anchor-
ing (about $1.007ft2); and research and development costs for de-
signing a  site-specific anchoring and  float system for the  cover
($2 to $3/ft2)." These costs would have been lower  if a less expen-
sive  cover material such  as  Hypalon ($0.707ft2) could have been
used and if  site access and berm stability  problems were not pres-
ent.  It is estimated that these factors almost doubled the cost of the
cover system for the Old Inger site.

Alternative Solutions for the Old Inger Site

  Other methods for controlling the potential  overflow from the
lagoon at Old Inger were evaluated.  Cover systems made from
steel, aluminum and wood were rejected because, at $10 to $20/ft',
these systems would exceed  the cost of pumping and treating the
lagoon wastes. Air supported structural covers were also rejected
on the basis of higher costs which would  exceed $8,00/ft' of cov-
ered area.22 Inflatible covers were rejected  because the lagoon berm
at the Old Inger site would probably not withstand the pressures
that  occur at the perimeters of these covers. Also, these covers re-
quire operation and maintenance of air pumps which would not be
convenient for the Old Inger site.
  Two additional alternatives considered  for the Old Inger situa-
tion  included  "no-action" and continued  periodic pumpdown/
treatment as required. The "no-action" alternative would require
no expenditures but would result in an increase in the level of con-
tamination  in the  swamp  surrounding the Old Inger site.  The
pumpdown/treatment  alternative  would   cost  approximately
$110,000 each time, assuming that 2 ft of liquid (i.e., approximate-
ly 540,000 gal) would be removed  at an average cost of $0.20/
gal.  If the  level  of the lagoon had not dropped during a recent
                                                        rain, and if 3 years were required for complete cleanup of the site,
                                                        it is estimated that the lagoon would require at least seven pump-
                                                        down/treatment  operations  for  an approximate  total cost of
                                                        $770,000.  This estimate is based on 60 in. of annual rainfall and
                                                        virtually no evapotranspiration from the lagoon due to the floating
                                                        oil emulsion layer. Therefore, a floating cover would have been less
                                                        expensive  (at  $6.5;/ft2,  the  total cost would  be approximately
                                                        $250,000)  than seven pumpdown/treatment operations at the Old
                                                        Inger site if the level of the lagoon had not dropped.
                                                        CONCLUSIONS

                                                           Presently, the most common method of preventing overtopping
                                                        of abandoned waste lagoons is pumpdown/treatment of the upper
                                                        2  ft of waste liquid.  Costing between $0.10 and S0.40/gal, this
                                                        method can be expensive, especially if it must be performed several
                                                        times prior to final cleanup of a site.' Floating covers can be a more
                                                        economical alternative; however, costs are extremely dependent on
                                                        site-specific factors. Covers generally cost SS.SO/ft'.1"20 Assuming
                                                        that the combined  cost of pumpdown, transportation and treat-
                                                        ment is $0.20/gal;  the  price for one  operation would be about
                                                        15% cheaper than a floating cover on  the same lagoon. This cost
                                                        relationship holds true for any size lagoon assuming vertical lagoon
                                                        walls and pumping 2 ft of impounded liquid. Therefore, if the com-
                                                        bined costs of pumpdown and treatment exceed S0.23/gal (i.e.,
                                                        $0.20/gal plus  15%),  or if pumping is required more  than once,
                                                        then a floating cover  may become the  most cost-effective method
                                                        of preventing overtopping of a lagoon. However, it is again empha-
                                                        sized that  adverse  site conditions may significantly increase the
                                                        unit cost of a floating cover.
                                                           The engineering considerations associated with site conditions at
                                                        the Old  Inger site almost doubled the normal  cost of a floating
                                                        cover. However, this cost  increase would not have caused the cost
                                                        of a floating cover to rise above the  costs  estimated  for pump-
                                                        down/treatment operations if the level  of liquid in the lagoon had
                                                        not dropped. The drop in lagoon liquid level at the Old Inger site
                                                        is  still under investigation. Unless site conditions change again,
                                                        neither a floating cover nor pumpdown/treatment operations will
                                                        be needed at the site.

                                                        REFERENCES

                                                         1. Tafuri,  A..  USEPA, OHMSB.  Personal communication with J.
                                                            Meade, JRB Associates, Dec. 1983.
                                                         2. Mason &  Hanger-Silas  Mason  Company, Lexington,  KY,  Wasie
                                                            Lagoon Spillage Control System, Phase 1, June 1983.
                                                         3. Brown, S.,  Burke Rubber Company, San Jose,  CA. Personal com-
                                                            munication with M. Evans, JRB Associates, Dec. 1983.
                                                         4. Burrows, W.C., Erosion Control Systems, Tuscoloosa, AL. Personal
                                                            communications with M.  Evans, JRB Associates, Dec. 1983.
                                                         5. Buichko, S.T., Gundle Lining Systems, Inc.. New York, NY. Per-
                                                            sonal communications with M.  Evans, JRB Associates, Dec. 1983 and
                                                            Jan. 1984.
                                                         6. Cain, G., Staff Industries Inc., Alexandria, VA. Personal commun-
                                                            ications with M. Evans, JRB Associates, Dec. 1983 and Jan. 1984.
                                                         7. Crumblis, R., Gulf  Seal Corporation,  Houston, TX. Personal com-
                                                            munications with M. Evans, JRB Associates, Dec. 1983.
                                                         8. Dupont Elastomer Information Center, Wilmington,  DE. Personal
                                                            communications with M.  Evans, JRB Associates, Dec. 1983.
                                                         9. Dyring, T., Schlegel Lining Technology, Inc., The Woodlands, TX.
                                                            Personal communications with M. Evans, JRB Associates, Dec. 1983.
                                                        10. Green, J., MWM Contracting Corporation, Auburn Heights, Ml.
                                                            Personal communications with  M.  Evans,  JRB Associates,  Dec.
                                                            1983.
                                                        11. Hardison, J.L., National Seal Company, Palatine, IL. Personal com-
                                                            munications with M. Evans, JRB Associates, Dec. 1983.
                                                        12. Kays, W.B., Globe Linings, Inc., Long Beach, CA. Personal commun-
                                                            ications with M. Evans, JRB Associates, Dec. 1983 and Jan. 1984.
                                                        13. Kinghorn,  R., Flexalon  Ltd.,  The Woodlands, TX. Personal com-
                                                            munications with M. Evans, JRB Associates, Dec. 1983.
410
ALTERNATIVE TECHNOLOGY

-------
14. Long, F., The BF Goodrich Company, Dallas, TX. Personal commun-
   ications with M. Evans, JRB Associates, Dec. 1983.

15. Moreland, J., MFC Containment Systems, Inc., Chicago,  IL. Per-
   sonal communications with M. Evans, JRB Associates, Dec. 1983.
16. Peloquin,  L., Sta-Flex Corporation, San Jose,  CA. Personal com-
   munications with M. Evans, JRB Associates, Jan. 1984.
17. Rodwin,  A., Watersaver Company, Inc., Denver, CO.  Personal
   communications with M. Evans, JRB Associates, Dec. 1983.
18. Sahol, J.R., Globe Linings, Inc., Houston, TX. Personal commun-
   ications with M. Evans, JRB Associates, Dec. 1983.
19. Shuey, R., Palco Linings, Inc., South Plainfield, NJ. Personal com-
   munications with  M. Evans, JRB Associates, Dec. 1983 and Jan.
   1984.
20. Way, W., Gulf Seal Corporation, Houston, TX. Personal commun-
   ications with M. Evans, JRB Associates, Jan. 1984.
21. Kays, W.B., "Roof System Controls Pollution in Reservoirs," Water
   andSewage Works, Nov. 1977.
22. Means, Building Construction Cost Data, 1984.
23. Denzer, D. Robert Snow Means Company, Kingston, MA.  Personal
   communications with M. Evans, JRB Associates, Dec. 1983.
24. Elwell, F.H., "Flexible Reservoir Covers: A Case Study," J AWWA.
    71, 1979, 210.
25. Solari,  K., USEPA, Region VI, Personal communications with M.
    Evans, JRB Associates, Dec. 1983.
26. Kittredge, D., "Floating Reservoir Covers for Manchester, N.H.,"
    Water and Sewage Works, Feb. 1971.
27. Bailey,  S.W. and Lippy,  E.G.,  "Should All Water Reservoirs Be
    Covered?" Public Works. 109,1978, 66.
28. Dial, et at., "Floating Cover for a Liquid Storage Reservoir." U.S.
    Patent 3,313,443, Apr. 1967.
29. Burke,  et al., "Reservoir Cover and Canalizing Means," U.S. Patent
    3,991,900, Nov. 1976.
30. Kays, W.B.,  "Gas Venting for Floating  Covers." U.S.  Patent
    3,980,199, Sept. 1976.
31. M. Putterman & Company, Inc., Chicago, IL, Personal commun-
    ications with M. Evans, JRB Associates, Dec. 1983.
32. Law  Engineering Testing  Company, Hydrogeologic Report and
    Development of Conceptual  Remedial Alternatives for  Old Inger
    Waste  Oil Refinery Site,  Submitted to Louisiana  Department of
    Natural Resources, Apr. 1982.
33. D'Appolonia/GDC. Final Report, Remedial  Investigation Phase 1,
    Element 1: Old  Inger Abandoned Hazardous Waste Site, Darrow,
    Louisiana. Submitted  to the Louisiana Department of Natural Re-
    sources, Oct. 1983.
                                                                                         ALTERNATIVE TECHNOLOGY
                                                                                                                                  411

-------
        EVALUATION  OF  ADVANCED TECHNOLOGIES FOR
                   TREATMENT  OF  CONTAMINATED SOILS

                                        WALTER  P. LAMBERT, Ph.D.
                                              LAWRENCE J. BOVE
                                               Roy  F. Weston,  Inc.
                                           West Chester, Pennsylvania
                                                 WAYNE E.  SISK
                              U.S.  Army Toxic and  Hazardous Materials Agency
                                     Aberdeen Proving Ground, Maryland
 INTRODUCTION
   Many parcels of Army real estate have been contaminated by ex-
 plosives, solvents  and  heavy metals  wastes that reached the soil
 from various industrial operations. Examples of these waste-gen-
 erating activities  included  equipment rebuilding and  repair,
 munitions manufacturing and munitions disposal.
   The Installation Restoration (IR) Program was established in the
 early 1970s to deal with potential soil contamination problems.  It
 was evident that many of the chemical  contaminants at Army in-
 stallations were either unique to the military or were predominant-
 ly associated with the military. Acceptable limits for soil residues
 often were not established. Those that were established were de-
 termined on a site-specific  basis through  negotiation with the
 appropriate regulatory agencies. Very few processes for removal,
 decomposition or immobilization of soil contaminants are avail-
 able or applicable to the Army's specific situations. A need for
 decontamination process development was recognized, and a num-
 ber of unit processes were investigated under the IR Program estab-
 lished to meet this need.
   The primary objective of the work reported in this paper was the
 identification of technologies: (1) not  previously investigated by the
 Army; (2)  which promised reasonable cost effective engineering
 solutions to specific types of soil contamination problems; and (3)
 which were potential subjects for further research and development
 within a set or predefined criteria.

 APPROACH TO TECHNOLOGY ASSESSMENT

   A structured technology evaluation  methodology was devised.
 Of necessity, it utilized limited information  and engineering judg-
 ment for execution. Two levels of evaluation were used. Each  con-
 tained a number of steps.

 Level 1—Conceptually Feasible Technologies

   The objective of Level 1 assessment was to filter out those tech-
 nologies which were conceptually applicable to decontamination of
 soils but which were not conceptually  feasible within  constraints
 peculiar to the Department of the Army.

 Step 1—Classification and Categorization of Soil
 Contamination Situations

   The objective of Step 1  was to define the array of soil contam-
 inants expected to be treated by whatever technologies appeared
 most promising. Since  few guidelines, criteria or regulations  were
 available for acceptable soil residual levels for the chemicals of in-
 terest, there were, by  definition, few  if any soil  contamination
 problems. Various situations did  exist,  however. They were iden-
                                                    tified by the actual or suspected presence of industrial chemicals
                                                    not otherwise found in the soil of a particular site or by the actual
                                                    presence of  naturally occurring materials at higher than back-
                                                    ground concentrations.
                                                      Those situations having available data were evaluated. Generic
                                                    types of soil contaminate profiles were developed. Classification
                                                    and categorization of situations were made by types of contami-
                                                    nants and by the association between types. It was found  that a pre-
                                                    dominant number of contaminated sites  contained solvents, ex-
                                                    plosives and heavy metals intermixed. This was the generic situa-
                                                    tion used for all subsequent evaluations.  Technologies showing
                                                    promise for treatment of onl> one or two types of contaminants
                                                    were identified separately and were not primary subjects for this
                                                    effort.
                                                    Step 2—Identification of Conceptually Applicable
                                                    Technologies

                                                      Unit processes or subsystems which appeared to be conceptually
                                                    applicable to the  treatment of one or more of the principal con-
                                                    taminant types were identified. This process produced an Initial
                                                    Technology List. The List was amplified by developing background
                                                    information and process descriptions for each entry. Technologies
                                                    were categorized as thermal,  chemical,  biological  or physical
                                                    processes. Published and unpublished information were used. In-
                                                    formation was  displayed  in a standard format to facilitate subse-
                                                    quent evaluation.

                                                    Step 3—Criteria Development
                                                      Level 1 assessment was conducted with a set of ten criteria char-
                                                    acterizing the feasibility of applying a particular technology. Each
                                                    criterion has three units of value ( + , 0, - ). A " + " was  favorable,
                                                    while a " - " was unfavorable. Specific definitions for each unit
                                                    were devised to guide the engineers and scientists making the tech-
                                                    nology assessments.

                                                      The criterion of technology performance was used to assess a
                                                    technology's ability to remove,  fix or decompose a specific soil
                                                    contaminant.  Technologies   with   demonstrated  effectiveness
                                                    against any  number of  the contaminants  of interest were rated
                                                    high. Technologies with  demonstrated ineffectiveness were rated
                                                    low. This criterion was  applied using best engineering judgment
                                                    since quantified performance standards were not available at the
                                                    time of the analysis.
                                                       + -Demonstrated decomposition or removal of identified soil
                                                         contaminants  (e.g.,  explosives-related  organics,  heavy
                                                         metals, solvents, PCBs and pesticides) to accepted levels
412
ALTERNATIVE TECHNOLOGY

-------
                             Table 1
         Thermal Processes Considered for Level 1 Assessment
                                                          FY84
Technology                                                Status
                                      Table 3
                  Biological Processes Considered for Level 1 Assessment
Vertical well chemical reactor
Multiple hearth incinerator
Rotary kiln incinerator
Molten salt incinerator
Fluidized bed incinerator
Wet air oxidation
In situ vitrification
Plasma arc torch
Microwave plasma detoxification
Burning/torching
Low temperature thermal decomposition
In situ hot air/steam stripping
In situ microwave heating
Laser-initiated thermal decomposition
Supercritical water oxidation
High temperature fluid wall reactor
AL
AL
SA
DV
AL
AL
DV
AL
DV
SA
DV
DV
DV
DV
DV
DV
AL = Alternate technology (in use for other purposes but conceptually adaptable to soil decon-
tamination).
DV = Developmental technology.
SA = State of the art technology.

                             Table 2
         Chemical Processes Considered for Level 1 Assessment
                                                          FY84
 Technology                                                Status
 Sulfur-based reduction
 Reduction with sodium borohydride
 On-site solvent extraction
 Solvent extraction-Acurex Process
 In situ solvent extraction
 Decontamination of soils using the Franklin Solvent
 Free radical oxidation
 Free radical oxidation-Enercol Oxidation Process
 Fenton's Reagent
 Base-initiated decomposition
 Carbon adsorption
 Ion exchange
 Surfactant complexing
 Complexing with dithiocarbamate
 Philadelphia Quartz (PQ) complexing agent
 Complexing with cellulose xanthate
AL
AL
AL
AL
DV
DV
AL
DV
AL
AL
AL
AL
AL
AL
AL
AL
 AL = Alternate technology (in use for other purposes but conceptually adaptable to soil decon-
 tamination).
 DV = Developmental.
 SA = State of the art.
   0   -Expected but not demonstrated  decomposition or  removal
      of identified soil contaminants
   -  -Minimal  demonstrated or  expected decomposition or  re-
      moval of identified soil contaminants
   The soils of interest had contamination profiles featuring both
 heavy metals and organic compounds.  The criterion of versatility
 was used to rate the ability of a technology to successfully remove
 both organics and inorganics. This criterion did not attempt to rate
 possible contaminant interferences to performance.
   +  -Demonstrated to decompose or remove both organic and in-
      organic contaminants of interest from soil
   0   -Expected but not demonstrated  decomposition or  removal
      of both organic and inorganic contaminants of interest
   -  -Expected or demonstrated ability to decompose or remove
      both organic and inorganic contaminants of interest  from
      soil
         Technology
                                                                                                                                FY84
                                                                                                                                Status
Microbial bioaccumulation of metals
Immobilized cells
Vermicomposting
Composting
Aerobic biodegradation
  -Activated sludge
  -Rotating biological contractor
  -Biopond
Biological fluidized bed
Landfarming
Anaerobic biodegradation
Anaerobic/aerobic cycling
Adapted microbial cultures
Vegetative uptake
Bioreclamation of soils (CDS System)
DV
AL
AL
AL

SA
SA
SA
AL
AL
AL
AL
DV
DV
SA
         AL = Alternate technology (in use for other purposes but conceptually adaptable to soil decon-
         tamination).
         DV = Developmental.
         SA = State of the art.
                                      Table 4
                  Physical Processes Considered for Level 1 Assessment
                                                                                                                                FY84
                                                                      Technology                                                Status
Secure landfill
Slurry wall
Grouting
Geological isolation
Stabilization (chemical admixing)
Microencapsulation
Macroencapsulation
High gradient magnetic separation
Washout
SA
SA
SA
DV
SA
SA
SA
AL
AL
AL = Alternate technology (in use for other purposes but conceptually adaptable to soil decon-
tamination).
DV = Developmental.
SA = State of the art.

  Sometimes  the  volume  of residuals produced by  a technology
is greater than the original contaminated volume, and a disposal
problem results. The criterion of volume of residuals was used to
rate a technology on its ability to reduce the volume of residual
material.
   +  -No additional residual volume generated
  0   -Residual volume significantly less than volumes of  treated
      soil
   —  -Residual volume equal to or  greater  than the volume of
      treated soil
  Soil treatment technologies, while removing contaminants from
soils, might produce  residuals that must be further processed. The
treatment or  disposal  of  residuals may be more costly than the
initial soil treatment. The criterion of a need for additional treat-
ment was used to score each technology on its ability to produce
nontoxic residuals. Systems that produced toxic residuals requiring
further treatment were rated low.
   +  -No additional treatment requirements
  0   -Additional  treatment for nontoxic or nonhazardous com-
      pounds required  (e.g., effluent gas scrubbing for particulate
      removal)
   -  -Additional  treatment for toxic or hazardous compounds re-
      quired
                                                                                          ALTERNATIVE TECHNOLOGY
                                                                      413

-------
  A decontamination process may alter the conditions of the site,
and this change may result in the movement of otherwise immobile
contaminants to other media. This movement could then have an
adverse effect on the local population, especially if groundwater
becomes contaminated. The criterion of intermediate transport was
used to rate technologies on their ability to minimize the spread of
soil contaminants to another environmental medium.
   + -Demonstrated minimization of transport of contaminants
     between environmental media (i.e., air, water and soil)
  0 -Unproven effects on intermedia transport
   - -Demonstrated assistance to the transport of contaminants
     between media
  It is possible that the operation of soil decontamination machin-
ery or the application of a specific unit process may create unsafe
conditions for the operators  or  for local residents. The safety
criterion was used to evaluate a technology on its inherent safety
implications for workers and local residents.
   + -No demonstrated hazard to workers or local residents during
     or after application
  0 -Hazardous materials contained or controlled with  minimum
     threat to workers or local residents  during or after applica-
     tion
   - -Demonstrated or strongly suspected hazards  to workers or
     local residents during or after application
  Since the volume of soils at contaminated sites is often very large,
a process should be able to treat contaminants at an efficient rate.
This criterion was used to rate each technology on its ability to de-
contaminate large volumes of soils within a reasonable time frame.
   + -Demonstrated ability to process large areas or large volumes
     of contaminated soil
  0 -Adaptability to process large areas or large volumes of con-
     taminated soil
   - -Demonstrated or expected limitations on the area or volume
     of soil that can be processed
  It is the goal of any decontamination  technology  to render a
once-contaminated site safe for unlimited  use. The  criterion of
future land use attempted to anticipate the results of  the applica-
tion of a technology to a given site. Sites that would be able to have
unlimited public  use upon completion of the treatment  process
rated high.
   + -Original contaminated site will have unrestricted public use
     after application
  0 -Original contaminated site will have limited public use after
     application (e.g., industrial use)
   - -Original contaminated site will be greatly restricted in use or
     access after application
  Contaminated sites of interest often contained both  organic and
inorganic contaminants. It is possible that a process that success-
fully removed organics  would not perform well in the presence of
inorganics and vice versa. This criterion was used to rate technolo-
gies on their ability to  perform in the presence of many contam-
inants.
  + -Demonstrated to be unaffected by the presence of other con-
     taminants of interest
  0  -Not expected to be affected by the presence of other contam-
     inants of interest
  - -Expected to be adversely affected by the presence of other
     contaminants of interest
  A technology which could be designed to be modular  or trailer-
mounted was considered very attractive.  Multiple  sites could be
serviced by the same hardware. A mobility criterion  was used to
rate a  system's potential  for  a modular,  transportable  design of
reasonable capacity.
  + -Demonstrated to be easily transported between sites
  0  -Expected to be adaptable for transport between sites
  - -Not transportable between sites

Step 4—Assessment Execution

  Ballots  were used  in the evaluation process. All technologies
conceptually applicable from the Initial Technology List were on
                                                        the ballot. One ballot was prepared for each of the ten feasibility
                                                        criteria. An assessment team comprised of chemical and environ-
                                                        mental engineers, hydrogeologists and chemists  was  assembled.
                                                        Each technology was presented to the assessment team twice: first,
                                                        using the written detailed technology descriptions developed in Step
                                                        2, and then verbally by the project engineers. The verbal presen-
                                                        tation was followed by discussions and completion of the assess-
                                                        ment ballots. Voters were encouraged to include explanatory com-
                                                        ments.
                                                          Ballots were collected and tabulated. Using  the majority vote,
                                                        one score (either + , 0, or - ) was selected for each criterion  for
                                                        each technology. Written comments were used to resolve ambig-
                                                        uous scores. Technologies receiving more pluses than minuses were
                                                        considered potentially or conceptually feasible. These technologies
                                                        were subsequently evaluated for their research and development
                                                        investment potential.
                                                             2 — Potential Research and Development
                                                        (R&D) Targets
                                                          The purposes of Level 2 assessment were to: (1) discriminate be-
                                                        tween technologies that were commercially available and those that
                                                        were developmental or conceptual at the time of the assessment,
                                                        and (2) identify technologies that were worth further research and
                                                        development investments.
                                                        Step 1— Criteria for R&D Investment Potential
                                                          Seven criteria were  developed. A value system similar to that
                                                        used for the Feasibility Criteria of Level 1 was used.
                                                          A critical programmatic goal under the IR Program was  to be
                                                        able to implement  new technologies no later than during Fiscal
                                                        Year 1987 (FY87). This criterion was used to differentiate between
                                                        those technologies that were commercially available at the time of
                                                        the assessment (FY84) from those which were anticipated or not an-
                                                        ticipated to be ready by FY87.
                                                           + -Commercialized or can be immediately adapted for removal
                                                             or decomposition of soil contaminants of interest
                                                          0  -Expected to  be commercialized  for removal or  decom-
                                                             position of soil contaminants of interest by FY87
                                                           - -Unlikely to be commercialized by FY87
                                                          The criterion of  proprietary status was  used  to  differentiate
                                                        between those technologies which were in the public domain (+)
                                                        and those for which there were important or  restricting proprietary
                                                        aspects ( — ).  The reason this criterion was included was because it
                                                        would affect  the ultimate cost of the technology developed.
                                                          +  -In the  public domain
                                                          0  -In the  public domain, but proprietary variations exist
                                                          -  -Proprietary
                                                          In general, all of the technologies identified as conceptually feas-
                                                        ible would require some level of R&D investment before they were
                                                        ready for full-scale  implementation.  A criterion based on the esti-
                                                        mated commercialization cost was used to evaluate the amount of
                                                        R&D investment required to bring a specific technology  from its
                                                        FY84 level of development to full-scale implementation.
                                                          +  -R&D costs are comfortably within the expected  budget
                                                          0  -Cost might stress the budget
                                                          -  -R&D costs expected to exceed reasonable budget constraints
                                                          Estimated  relative operational costs were difficult to quantify at
                                                        the time of the assessment because there was insufficient informa-
                                                        tion available. This  criterion utilized the  rating team's best esti-
                                                        mate of operational costs using a secure  landfill  as  the basis for
                                                        comparison.
                                                          +  -Less than disposal in an off-site secure landfill
                                                          0  -Approximately equal to disposal in an off-site secure landfill
                                                          -  -Significantly greater than disposal in an off-site secure land-
                                                             fill
                                                          The application  of a technology to a contaminated site will al-
                                                        ways have some affect on the area's ecology. A criterion based on
                                                        environmental impact  was used to assess the ability of a contam-
                                                        inated site to recover after application of a specific remedial action
                                                        technology.
414
ALTERNATIVE TECHNOLOGY

-------
  -i-  -impact during application is insignificant or within reason-
     able limits,  with  complete ecological recovery after applica-
     tion
  0  -Adverse impact during application with full ecological recov-
     ery expected after application
  •-  -Significant adverse  impact  during application with  only
     partial ecological recovery expected after application
  Many times,  removal technologies are contaminant-specific. A
cleanup technology may be able to remove only a select group of
solvents, for example. Given the contamination profiles anticipated
at various Army installations, it might be necessary to employ sev-
eral technologies to remove all contaminants at a specific site. This
criterion assessed the ability of each process to either stand alone or
be easily linked with other unit processes.
   + -Either stand-alone or easily linked with other unit processes
  0  -Unknown capability to be linked with other unit processes
   - -Cannot easily be linked with other unit processes for devel-
     opment of a system; linkable only in disjointed or discontin-
     uous systems
  Versatility was a criterion also used in the Level 1 assessment.
It was used in the Level 2 assessment to further point out the ability
of technologies to remove both organics and inorganics from con-
taminated soils.
  Four questions on topics warranting further consideration were
included in addition to the set of seven evaluation criteria. The
questions were not discriminatory in the sense that  technologies
would survive or fail Level 2 assessment, but they were felt to be
relevant to final recommendations on R&D investments.
•Is process equipment, hardware, software or a demonstration sys-
  tem currently available for field demonstration or pilot-scale test-
  ing? (Yes/No)
•Can reliability be designed into the hardware/software subsystems
  without exorbitant expense? (Yes/No)
•Are exotic or strategic materials required? (Yes/No)
•Are there materials handling risks to be assessed?  (Yes/No)
Step 2—Level 2 Assessment Execution
   Ballots were again  used in the evaluation process. The  same
assessment team that evaluated the conceptually  feasible technol-
ogies was used in this second evaluation process. Ballots were pre-
pared for each  of the technologies.  Each ballot contained the seven
assessment criteria and the four nondiscriminatory questions de-
tailed under Step 1.
   The time required to bring a given technology to full-scale im-
plementation  was the  most  important discriminator in  Level 2
assessment. Technologies or unit processes which required  more
than four calendar years  for commercialization were eliminated
from further consideration. Technologies meeting the  time to com-
mercialization criterion were given further consideration. Ultimate-
ly, technologies that could be fielded within  a minimum timeframe
(fast track) and those  which could be fielded within the four-year
time constraint (medium track) were sought.
   Level 2 assessment was  made on the basis of information used
for  Level 1 and new  information obtained between evaluations.
Conflicts in scores were resolved  by referring to evaluator  com-
ments accompanying each rating.

RESULTS OF TECHNOLOGY ASSESSMENTS
Initial Technology List

  The technologies which were judged to  be conceptually  appli-
cable to the removal, fixation or decomposition of military unique
or relevant soil contaminants are listed in Tables 1  to 4. For further
information, the reader is referred to the report of Bove, et al.'
  Of the 55 technologies listed, 25  were judged to be  conceptually
feasible  for consideration  as potential  R&D investment  targets
(TableS).
  Fast track and medium track technologies were the first results
of Level 2 assessments (Table 6). Fast track technologies were those
which were judged  to require the minimum  time to full-scale util-
ization. Medium track technologies  were judged to be ready for
                             Table 5
Technologies Surviving Level 1 Assessment and Judged to be Conceptually
    Feasible for Removal of Military Unique or Military Relevant Soil
                          Contaminants
  In situ vitrification
  Rotary kiln incinerator
  Low temperature thermal
  Plasma arc torch
  Fluidized bed incineration
  Solvent extraction (Acurex)
  Free radical oxidation
  Franklin Solvent
  Fenton Reagent


  Composting
  Landfarming


  Secure landfill
  Macroencapsulation
  Microencapsulation
  Thermal Processes
High temperature fluid wall reactor
Microwave plasma
Multiple hearth incinerator
Super critical water oxidation

  Chemical Processes
On-site solvent extraction
In-situ solvent extraction
Surfactant complexing
On-site base-initiated reduction
  Biological Processes
Vermicomposting

  Physical Processes
Geological isolation
Magnetic separation
                             Table 6
    Preliminary Level 2 Assessment Results for Fast and Medium Track
             Potential Research and Development Targets

                    Fast Track Developmental Targets
  Secure landfill                          Rotary kiln incinerator
  Multiple hearth incinerator                 Microencapsulation
  Macroencapsulation                      Landfarming
  Fluidized bed incineration                  Composting
  Vermicomposting
              Medium Track Research and Development Targets
  Geological isolation                      On-site solvent extraction
  Surfactant complexing                    Fenton reagent
  Base-initiated reduction                    Free radical oxidation
  In situ solvent extraction                   Low Temperature thermal
  Acurex Process                         In situ vitrification
  High temperature fluid wall                 Microwave plasma
  High gradient magnetic                    Plasma arc torch
  Franklin Solvent
full-scale application within the four-year time constraint, but they
required more testing and investment than the fast track technolo-
gies.
  Whether the Army invests R&D funds in one or more of the tech-
nologies listed in Table 6, depends upon a number of factors such
as redundancy between candidate topics, acceptability to Depart-
ment of the Army from a regulatory viewpoint, projects in pro-
gress and  work planned or in progress by other government agen-
cies.

ACKNOWLEDGMENT AND DISCLAIMER

  The bulk of the work reported in this paper was funded by the
U.S. Army Toxic and Hazardous Materials Agency as part of an
ongoing program to develop new or adapt existing technologies to
the remediation  of actual or potential soil contamination problems
on Department of Defense installations. The work represents activ-
ities and investigations conducted during an intermediate stage in
the program.  The information displayed in this paper represents
the opinions and views of the  authors  and does not constitute
official views or policies of the Department of the Army, the De-
partment of Defense or any other Federal agency, office or person.

REFERENCES

1. Bove, L.J., Cundall, C.L., Lambert, W.P., Marks, P.J. and Martino,
  J.F., Removal of Contaminants from  Soil. Phase I: Identification
  and Evaluation of Technologies. Final  Report No.  DRXTH-TE-CR-
  83249 for the  U.S.  Army  Toxic and Hazardous Materials Agency,
  Aberdeen Proving Ground, MD,  under Contract DAAKll-82-C-OOn'
  Dec. 1983.
                                                                                       ALTERNATIVE TECHNOLOGY
                                                            415

-------
    ALTERNATIVES  FOR DISPOSAL  OF  UNKNOWN  GASES,
               CHEMICAL CONTROL  CORPORATION SITE,
                                ELIZABETH,  NEW JERSEY
                                                DAN NICKENS
                                                  JEFF GOLD
                                      Earth  Resources Consultants, Inc.
                                                Orlando,  Florida
                                                HENRY MUNOZ
                                               Corps of Engineers
INTRODUCTION

  Compressed gas cylinders containing unknown or highly toxic
material are being discovered at an ever increasing rate. They have
been found at  waste disposal sites, research facilities,  landfills,
military installations and a variety of other locations. Although the
total number of this type of cylinder is not high, the potential for
an uncontrolled release of highly toxic gases poses significant prob-
lems.
  The  Chemical Control Corporation  Superfund  site  in Eliza-
beth, New Jersey is an example of a site where waste gas cylinders
pose continuing disposal  problems. The initial remedial actions
taken at the site were confounded by the discovery of compressed
gas cylinders with  unknown contents. Efforts by the State of New
Jersey,  the USEPA, the U.S. Army Corps of Engineers and pri-
vate contractors to permanently dispose of the cylinders have been
unsuccessful to date.
  One-hundred and eighty unidentifiable compressed gas cylinders
in varying degrees of deterioration were discovered following the
1980 fire at the Chemical Control Corporation site. The variety of
cylinder types present at the Chemical Control site ranges from
aerosol cans and lecture bottles to large propane-type tanks. The
cylinders lack any  identifying markings, are in poor structural con-
dition and typify the kind of cylinder which can pose serious prob-
lems to those faced with disposal responsibilities.
  Transportation  of unknown materials over public  thorough-
fares is prohibited by Federal law for interstate  movements and
by most states for intrastate shipments. To transport cylinders con-
taining  unknown gases  for the purpose of disposal by detonation
or any other means involves obtaining special transport permits and
emergency waivers. In addition, special precautions such as police
escort, coordination with local emergency services (police, fire de-
partments and civil defense) and transport along prescribed routes
during low-flow traffic periods are generally required.
  In developing the specifications for remedial action at the Chem-
ical Control Corporation site, a number of different options were
examined for disposal of the compressed gas cylinders. A majority
of these options were eliminated  by the unknown nature of the
cylinder contents.  Without exception, commercial  facilities special-
izing in gas cylinder disposal refused to accept the gases without
prior knowledge of their exact nature. Military decommissioning
facilities would  not accept the material in accordance with Depart-
ment of Defense policy and directives regarding the acceptance of
non-military hazardous materials for processing.  On-site disposal
schemes have also been dismissed as too risky considering the busi-
nesses and residences near the site. Given these limitations, it be-
came evident that sampling and analysis of the cylinder contents
                                                   was necessary prior to disposal of the cylinders and any gas they
                                                   might contain.
                                                   Sampling and Analysis
                                                     Sampling the cylinders by simpK opening the vahes (if this was
                                                   even physically possible) was dismissed because an uncontrollable
                                                   release leading to significant environmental contamination could
                                                   occur if the valve should break internally, break off at the neck of
                                                   the cylinder or open and not  close again.  Two potentially accep-
                                                   table  mechanisms for accomplishing this  type of sampling were
                                                   available:  (1) Cold tapping of a new valve into the side of the
                                                   cylinder and (2) Utilization of a cylinder rupturing device in a con-
                                                   trolled environment.
                                                     Cold tapping was deemed too dangerous due to inherent dangers
                                                   in the technique  and  the deteriorated condition of the cylinders.
                                                   The technique of sampling the cylinders using a cylinder rupture
                                                   unit is still under  consideration as the technology is currently being
                                                   developed.
                                                     In some cases, one may speculate on the contents of cylinders
                                                   after examining the type of cylinder and valving (e.g., port size and
                                                   thread arrangement).  The problem with this approach is that the
                                                   cylinder often is  not holding the type of gas for which it  was in-
                                                   tended; a  cylinder may  have been used as a receptacle for other
                                                   gases or the cylinder may be a "hybrid" made up of parts and com-
                                                   ponents from different cylinders. Efforts to identify the cylinders
                                                   and their contents based on visual inspection were unsuccessful due
                                                   to the degraded  condition of the cylinders after exposure to the
                                                   original fire and subsequent weathering.
                                                     The principal problems in handling these cylinders stem from:
                                                   •Deteriorated physical condition
                                                   •Non-functional or damaged valve mechanisms
                                                   •Obliterated identification markings
                                                   •Evidence of non-specified refilling (i.e.,  type A  gas placed into
                                                     type B cylinder)
                                                     Any one factor or a  combination of factors can pose serious
                                                   problems  in sampling and identifying the cylinder contents. Iden-
                                                   tification  of the  cylinder contents is a critical step  for  handling,
                                                   transportation and proper disposal.

                                                   CYLINDER DISPOSAL ALTERNATIVES
                                                     The proper disposal of waste cylinder contents  depends on the
                                                   nature of the waste. Many different alternatives are available for
                                                   cylinder disposal  after determining the nature of the contents.
                                                     Commercial facilities are readily available for handling various
                                                   contents via either incineration of pyrophoric gases and liquids or
                                                   chemical  treatment/neutralization of non-pyrophoric  contents.
416
ALTERNATIVE TECHNOLOGY

-------
 Some non-hazardous gases may be simply vented to the atmos-
 phere. Each of these alternatives, however, requires that the nature
 of the cylinder contents be known for safe and effective handling.
 Venting

   Various gases such as nitrogen and helium may be safely released
 to the environment  merely by releasing the contents through the
 valving mechanism or via cylinder rupture. This option is simple,
 inexpensive and the technology is readily available to open or rup-
 ture the cylinders.
   Prior to any venting operation, however, the nature of the cylin-
 der contents must be known. A great many industrial and  research
 gases have hazardous characteristics which would make  their re-
 lease catastrophic. The consequences of such a release would  be
 particularly severe in a heavily populated area such as the  town of
 Elizabeth, NJ.
 Incineration

   Incineration is a technology which can be employed for  disposal
 of flammable gases  and  various biological materials.  Commercial
 incineration facilities capable of handling many of the flammable
 gases are readily available and costs are minimal  for this type of
 disposal.
   Incineration disposal  technology may only  be employed for
 pyrophoric gases (such as acetylene and hydrogen sulfide), liquids
 and certain biological materials. Again, it  is  essential  that the
 nature of-4hewaste be  known,  this time in order to determine
 the  effectiveness^ol  the combustion process. Even among various
 flammable gases, the exact nature of the combustion process is de-
pendent upon the waste type as gases  are destroyed at different
 temperatures, some react explosively and some produce hazardous
 byproducts upon heating and combustion.
   Incineration is not a viable disposal alternative for the cylinders
 located at the Chemical Control Corporation site without sampling
 and analysis of their contents. In addition, commercial  inciner-
 ators will not accept pressurized containers due to the potential for
 explosion of a cylinder  upon heating.  Every incineration facility
 contacted declined acceptance of these cylinders, citing concerns
 over potential liabilities, safety  and permit restrictions:  (1) the
 cylinders contained unknown gases and (2) processing pressurized
 containers in an incinerator poses difficulties.
 Chemical Treatment

   Chemical treatment or neutralization can effectively  be em-
 ployed for most compounds and gases. Commercial facilities with a
 wide variety of treatment processes are available.  At least one
 facility capable  of treating industrial gases is located in  the geo-
 graphical area of the Chemical Control site; several other  sites are
 located across the country.
   Like incineration, effective chemical treatment or neutraliza-
 tion requires that the nature of the waste requiring treatment be
 known. A single process or even series  of processes cannot insure
 that all types of wastes will be effectively handled. For each waste,
 there is a specific process which will most effectively treat or neu-
 tralize that particular chemical compound.
   The potential for  a wide range of toxic wastes to be present at
 the  Chemical Control site eliminates  the utilization of  existing
 commercial facilities for  processing these cylinders. All commercial
 facilities  contacted about  the  cylinders declined  to accept them
 prior to sampling and analysis.
   On-site chemical  treatment is  similarly impractical because of
 the  wide variety of wastes potentially present. It would not  be
 possible to devise a  process or series of processes  capable of han-
 dling the range of potentially present  compounds. Additionally,
 construction of on-site  treatment facilities would be costly and
 could not insure against environmental releases.

 Ocean Disposal

   The USEPA has approved the disposal of certain wastes at off-
 shore locations. A disposal site which is approved for certain types
of waste is located at the 106-mile marker off the Atlantic coast.
This disposal alternative would require recontainerization of the
cylinders and shipment to this location.
  Ocean disposal of the cylinders would provide a simple and
fairly inexpensive means of disposal. The chemical Control site
which is located adjacent to a waterway would be easily accessible
by barge. Overland transport  of the cylinders (and the associated
risk  of  exposure along the  route) could be eliminated using this
option.
  This disposal alternative presents potentially great environmental
risks. Many gases commonly used  in industrial and research appli-
cations  are reactive in water.  The incompatibility of these wastes
with other wastes disposed of at this  site is also a potential prob-
lem. Of far greater concern is the potential for release of biolog-
ical or virological agents from  the cylinders into the ocean environ-
ment. For these reasons, the  USEPA representatives have stated
that under no circumstances would the ocean dumping of any un-
known wastes be approved.
On-site  Detonation

  On-site detonation at the Chemical Control site  represents the
most direct  means  for disposal of the compressed gas cylinders.
This could be implemented by evacuating the nearby populated
areas and setting shaped explosive charges on each cylinder.
  The explosive release of unknown gases could pose extreme haz-
ards to  the environment and any nearby populated areas. It is im-
possible to predict the exact  consequences  of the release of un-
known  gases or liquids. While construction of various  types  of
containment structures would help minimize release of contam-
inants and provide some control  over the release, the  ultimate
effect would be to disperse  the contents into the environment at
the site.
  Detonation of non-pyrophoric gases such as phosgene and boron
trifluoride could result in spreading these gases over a wide area.
Gases such as phosgene and chlorine, which are heavier than air,
tend to  accumulate in low-lying areas and could potentially pose
significant health hazards upon release of a  sufficient quantity.
These potential  hazards prevent on-site detonation from being a
viable alternative.

Off-site Detonation

   Off-site  detonation represents  the only presently existing and
tested disposal technology for handling gas cylinders of unknown
contents. In the past, this was the  principal method employed for
disposing of this type of cylinder.  The technology for imple-
menting this alternative is readily available and has been emplolyed
in the past for disposing of known gases contained in cylinders.
With the  increase in  the number  of  cylinders requiring disposal
plus greater environmental  and safety awareness,  however, this
option is no longer acceptable. During the initial remedial action
at the   Chemical Control site,  administrative project personnel
proposed detonation of the  cylinders  at a remote location in cen-
tral New Jersey. This plan was subsequently rejected, however, in
view of  health and environmental considerations.
  The detonation process can be used in conjunction with other
techniques to minimize the risks associated with this type of dis-
posal. A bunker may be constructed to contain the explosions,
reduce  the  noise and  limit  the  spread  of  contaminants.  The
detonation can also be used to simultaneously ignite gasoline and
other combustibles to aid in  the destruction  of certain  types  of
flammable gases. The technique can be varied slightly to provide
for a non-explosive rupturing of the cylinder structure with subse-
quent ignition in the detonation structure.
  The greatest disadvantage of using  the detonation and pyrolysis
techniques for cylinder disposal is the incompatibility  of  many
common gases with this disposal  method. Flammable gases con-
stitute only a portion of the  hazardous gases and wastes poten-
tially present in the cylinders. Many of the more toxic gases would
be completely unaffected by the technique and could be uncon-
trollably spread into  the environment. Non-flammable, liquified
                                                                                      ALTERNATIVE TECHNOLOGY      417

-------
gas could also be  released over large areas  through explosive
releases. Many toxic virological agents would be unaffected by  the
fire and could be released by the detonation. Even the flammable
gases could produce toxic combustion by-products or react syner-
gistically upon detonation to produce a much  more powerful ex-
plosion.
  Although the risks to populated areas can be  minimized through
selection of remote locations for  detonation, the  environment at
the disposal site  may become contaminated.  This could require
some type of remedial action and expose workers to hazardous
conditions.
  The viability of air  monitoring  which could  be conducted on a
real-time  basis during a detonation operation  is questionable.
Quantified data of the releases could only be  obtained at a later
date through sampling and laboratory analysis,  thereby minimizing
its usefulness in protecting both personnel and the environment.
  No commercial facilities are  available or permitted for opera-
tions of this type. Disposal facilities contacted  in this regard indi-
cated that the cylinders could not be accepted due to permitting
considerations and liability concerns. Past incidents of detonation
have required special waivers and have been conducted on a one-
time only basis at remote locations.  Commercial  firms are avail-
able to provide the detonation service prolvided  the client can furn-
ish a site. For these,  Governmental  facilities were considered as
potential detonation disposal sites.
  As a rule, Government facilities will not accept waste materials
from non-Governmental sources for disposal.  This directive may
be waived in emergency  cases when  public health and safety  are
being threatened. In most cases, however, disposal by detonation
or  any other means on a Government facility  should not be con-
sidered a viable option.

MILITARY OPTIONS
  The Department of Defense (DOD) currently operates a  gas
cylinder decommissioning facility at  Tooele Army base near Salt
Lake City, Utah. This facility utilizes remote handling apparatus to
unscrew  the valves from a cylinder and release the gaseous con-
tents into high temperature incinerators. The body of the cylinder
is then heated to drive off any remaining material and the emptied
cylinder is disposed of properly. Another facility of this same type
is under construction  at  a DOD facility in the Pacific. These  de-
commissioning facilities  handle only DOD materials (principally
gaseous warfare  agents) and, in all  cases,  the  contents of  the
cylinders are known prior to disposal.
  Experimentation  is  currently under way utilizing low tempera-
ture liquified hydrogen and nitrogen to super-cool gas cylinders.
The cylinders are essentially "frozen" and then  crushed. The solid-
ified gas and the cylinder body are  then incinerated under high
temperature.
  The DOD also  operates munition test ranges and artillery prac-
tice areas  that could potentially be used in the disposal scheme for
the cylinders.  In  accordance with government  policies, however,
these operational centers are restricted to  military use only.  The
only exception to this directive would be in cases where there is an
imminent threat to public health and safety. This restriction severe-
ly limits the usefulness of the military as a disposal option for non-
military wastes.

CYLINDER SAMPLING ALTERNATIVES
  Sampling of compressed gas cylinders is a risky procedure even
under ideal circumstances. Several deaths and many injuries have
resulted from attempts to sample cylinders even  where the contents
were known. The problems of sampling are considerably greater at
the Chemical Control site where the cylinders  are in a degraded
condition  and the contents are unknown.

Controlled Venting
  In most cases,  gas  cylinders  can easily be  sampled  using  the
valves on the cylinders. The contents can be slowly released into a
sample container and the cylinder re-sealed after sampling. This
                                                        sampling technique utilizes readily available technology and is in-
                                                        expensive.
                                                          Sampling  through the cylinder  valves on older or degraded
                                                        cylinders is a hazardous operation because of the potential for fail-
                                                        ure of the valve to  re-seal. This technique may only be utilized
                                                        where the valve mechanism is intact and in  good operating con-
                                                        dition.
                                                          The cylinders at the Chemical Control site are deteriorated to the
                                                        point  where this sampling technique is impractical. Most of the
                                                        valve mechanisms were severely damaged by the explosion and fire
                                                        and the  subsequent  weathering. The valve mechanisms which re-
                                                        main relatively intact exhibit evidence of weathering which renders
                                                        their functional viability questionable.
                                                          Potential exposure of the local population to toxic gases would
                                                        be greatly increased during sampling  activity using the  existing
                                                        cylinder  valves. The  risk of an accidental release is high unless the
                                                        sampling operation is carried out in a contained environment. This
                                                        technique could best be employed at a remote location where the
                                                        only exposure would be  to properly protected handling personnel
                                                        and the local environment. Obtaining an appropriate site for such
                                                        an operation would remain a logistical problem.
                                                          In certain cases, a cylinder can be safely sampled by attaching a
                                                        second valving component to existing connections on the cylinder
                                                        valve.  This can only be accomplished in situations where the pri-
                                                        mary valve housing  is  in good condition. This technique  (termed
                                                        the "Line-Purge" method) permits relatively safe and inexpensive
                                                        cylinder sampling where it can be applied. The Line-Purge method
                                                        would be impractical and  unsafe, however, for  most of the
                                                        cylinders at the Chemical Control site.
                                                        Cold-Tapping of Cylinders

                                                          Cold-tapping is a  technique which has been utilized to install
                                                        valves and other connections to  tanks and water lines. This tech-
                                                        nique has not, however, been successfully applied to high-pressure
                                                        compressed gas or  liquid structures.  The pressures  commonly
                                                        found in  compressed gas cylinders are incompatible with  existing
                                                        cold-tap  technology.  Cold-tapping of the cylinders at the Chemical
                                                        Control  site, especially in light  of the degraded  structural con-
                                                        dition, would be extremely hazardous and impractical.
                                                        Cylinder Rupture Vessel

                                                          The design of a structurally secure and mechanically ooerable
                                                        vessel  for remote rupturing and recontainerizing of gas cylinders
                                                        has been patented and is in the preliminary phases of fabrication.
                                                        This device, termed the Cylinder Rupture Vessel (CRV), is present-
                                                        ly being  fabricated by  a  private corporation. The CRV unit con-
                                                        sists of a pressure vessel which encloses the cylinders as they are
                                                        ruptured in an inert  atmosphere by means of a hydraulic punch.
                                                        After the cylinder contents have been released, the gaseous and/or
                                                        liquid  phases are  sampled and  the residue  recontainerized into
                                                        secure containers. The entire operation is accomplished remotely in
                                                        a secure chamber which provides a fail-safe containment system.
                                                          The risk of exposure to both operating personnel and local pop-
                                                        ulations  would be virtually non-existent during  CRV sampling
                                                        operations. The CRV is  a completely mobile, self-contained unit
                                                        and is able to process cylinders ranging from 1  to 24 in. in diameter.
                                                        This device, when constructed and fully tested, will provide a safe
                                                        alternative for sampling compressed gas cylinders.

                                                        TEMPORARY MITIGATION MEASURES

                                                          In order to minimize the risk of an uncontrolled leak of poten-
                                                        tially toxic gas, the 180 gas cylinders at the Chemical Control site
                                                        were overpacked  into specially fabricated recontainers in July of
                                                        1984.  The recontainers  were  designed,  fabricated  and certified
                                                        under  the ASTM pressure vessel code and consisted of a flanged
                                                        cylinder, closed at one end by a hemispherical end cap. The blind
                                                        flange top was fitted with a bleed-off valve and a pressure gauge
                                                        to monitor any leakage of the gas cylinder after it was placed into
                                                        the recontainer. A fiat plate was attached to  the end cap to allow
                                                        the recontainer to be  placed in an upright position for storage.
418
ALTERNATIVE TECHNOLOGY

-------
  Use of specialized recontainers on deteriorated gas cylinders is
acceptable as a temporary measure that should be used only to
eliminate an imminent danger to public health and safety. The
technology cannot be applied in lieu of final  disposal since the
problem cylinder has not been destroyed but merely temporarily
contained.
CONCLUSIONS

  Sampling, transporting and disposing of compressed gas cylin-
ders containing unknown gases pose  a  significant  problem.
Cylinders are being discovered in increasing numbers as the na-
tional effort to cleanup toxic waste sites proceeds. In most cases,
the problem is not as severe as at the Chemical Control site with 180
cylinders.  It is the  potential  toxicity, however,  rather  than the
quantity of cylinders at a site which constitutes the true hazard.
  Options for disposal of compressed gas cylinders whose  con-
tents are not known are extremely limited. Uncontrolled release
through rupture or detonation at remote locations has become an
unacceptable option from both an environmental and public health
standpoint. Commercial disposal facilities are available for gaseous
compounds, but the gas being destroyed has to be identified prior
to disposal as a prerequisite. Military decommissioning facilities
exist but are unavailable to process any non-DOD waste material
and are set up to handle only very specific waste materials.
  The  emphasis in  handling unknown compressed gas cylinders
has currently shifted to sampling the container contents so that the
proper disposal method can be employed. Existing technologies
for sampling gas  cylinders, especially those in deteriorated  con-
dition, are inadequate. New methodologies, such as the CRV, and
the Line-Purge method offer the most promising solutions to the
problems of properly handling and disposing of potentially  haz-
ardous gases contained in unmarked or unidentifiable cylinders.
                                                                                    ALTERNATIVE TECHNOLOGY      419

-------
DETOXIFICATION  OF SOILS, WATER AND BURN  RESIDUES
          FROM A MAJOR AGRICULTURAL  CHEMICAL
                              WAREHOUSE FIRE

                                  MARK D. RYCKMAN
                     Ryckman's Emergency Action and Consulting Team
                                   St. Louis, Missouri
INTRODUCTION

  At 5:00 pm on Apr. 23, 1980, a Tire broke out at the Hewitt-Ware
Feed and Supply Company warehouse in Hillsboro, Illinois (Fig.
1). Because of the potential acute and chronic life safety threats and
environmental impacts, local authorities contacted REACT for
assistance in responding to this emergency. REACT was asked to
provide safety recommendations, including designation of toxic
corridors for evacuation; the Firm was also requested to design and
                                       implement a comprehensive decontamination/remedial action pro-
                                       gram.
                                       RISK ASSESSMENT AND
                                       IMMEDIATE RESPONSE

                                        REACT engineers and scientists conducted an on-fcene rak
                                       assessment which revealed that people downwind from the fire
                                         Figure I
          Hewitt-Ware Feed and Supply Company. Undamaged Grain Elevator and Burned Agricultural Chemical Warehouse.
420
CASE HISTORIES

-------
should be evacuated because of the potential formation of toxic
gases  including cyanide, phosgene,  chlorine, fluorine,  oxides of
nitrogen and isocyanates.
  Four hundred families were evacuated from a 6,000-ft toxic cor-
ridor formed downwind from the fire. Hillsboro schools were clos-
ed for the day. A hospital and nursing home located approximately
2,000 ft downwind from the burn-site were not  evacuated due to
potential risks of moving elderly people and hospitalized patients.
However, all  intake systems,  windows  and air vents in both
faculties were closed during the fire to minimize the influx of toxic
gases. Because of the proximity of these facilities to the fire,  water
was applied to the fire in an attempt to "knock down" toxic fumes
and reduce the amount of toxic air pollutants in the cloud of smoke
generated by the fire.
                           Figure 2
       Containment Pits 1 and 2 Collect Concentrated Pesticide
               Runoff from Firefighting Operations.
  Hillsboro City personnel took quick action to block drainage
channels from the area; they also dug pits to minimize runoff of an
estimated 250,000 gal of water used to extinguish the fire (Fig. 2).
This quick action substantially reduced the water pollution poten-
tial from the highly contaminated water runoff. Using REACT's
computerized toxic corridor projection  system, evacuation limits
were reduced from 6,000 ft to 2,000 ft 6 hr after the fire began, and
finally to 800 ft 12 hr after the beginning."
  After the fire was extinguished, it was determined that a massive
hazardous  material cleanup program was required  to eliminate
public health hazard  from the remaining unburned pesticides and
the contaminated water.
  Fire fighting  operations had  caused a water main between
Schram City and Hillsboro to collapse. Two valve pits adjacent to
the fire filled up with contaminated runoff. This situation posed an
imminent direct cross connection problem with the drinking water
systems of both Schram City and Hillsboro. To mitigate this prob-
lem, the valve pits were immediately drained, the distribution lines
were  back flushed and service was restored.  City workers con-
structed a temporary dam to prevent contamination of downstream
water  supplies,  a potential  fish  kill  and  threat  to  livestock.
However, 300,000 gal of surface waters of Shoal Creek were con-
taminated above the dam.

ISOLATION AND CONTAINMENT
  Contaminated stream waters were isolated and transferred into a
300,000 gal polyethylene-lined  lagoon  (Fig. 3). The interceptor
trench collection network at the site of the fire was lined with ben-
tonite/sand bags and activated carbon. Water in the primary col-
lection pit at the site was pumped in to  55-gal  drums pending the
results of on-going treatability investigations (Fig. 4). Swales were
constructed to divert the uncontaminated portion of the watershed
around the contaminated area. Due to impending thunderstorms, a
plastic canopy was used to cover the entire burn site to prevent con-
taminant transport off-site.
TREATMENT AND DISPOSAL

  Twenty-one different pesticides, including 11,000 Ib of granular
chemicals and 3,465  gal of liquid chemicals, were involved in the
fire. Since these products had been exposed to  high temperatures,
the  exact  physical/chemical and toxicological  properties  were
unknown.  The known quantities and properties of  the material
present before the fire are presented in Table 1.
  The presence  of  numerous  known and  unknown complex
organics required the pragmatic selection of physical/chemical and
biological  indicator  decontamination yardsticks. Five pesticides
were  selected (Table 2), based on their relatively high aquatic and
mammalian toxicities  and water solubilities. Gas chromatographic
analyses were performed on samples extracted from the fire debris
on  each  of the five selected pesticides.  Sample spikes  were con-
ducted along  with the  evaluation  of  samples to  provide good
laboratory quality control.
  Aquatic bioassays were performed to  determine the presence of
any unknown  contaminants and to confirm detoxification opera-
tions. Five waste streams were identified, and alternative recovery,
treatment and disposal methodologies were considered. Ultimate
treatment criteria were based on bench-scale tests; key  to the
analyses were  treatment efficiency and cost. Pesticide detoxifica-
tion alternatives included: aeration, evaporation, solar  oxidation,
photolysis,  carbon  adsorption,  alkaline hydrolysis,  chemical
precipitation and oxidation with hydrogen peroxide. Five waste
streams were identified for treatment: contaminated stream waters;
concentrated leachate at the site;  soil from the site; contaminated
and uncontaminated building materials; and contaminated product
residues.
Contaminated Stream Waters

  A pumping  station was constructed to transfer the contents of
the stream into a 300,000 gal treatment lagoon constructed adja-
cent to the stream. This lagoon functioned as  a batch reactor for
                                                                                                  CASE HISTORIES
                                                          421

-------
                                                            Figure 3
                                            300,000 Gallon Treatment Lagoon Containing
                                              Contaminated Runofff from the Creek.
                                     The lagoon was lined with 24 mil Visqueen with side slopes of
                                    1.5 to 1. Pumping station transferring creek to treatment lagoon.
                                                    final grade lagoon apron.
the  following  treatment  operations:  aeration,  solar  oxidation,
evaporation, adsorption and clarification. Circulating pumps were
utilized in the lagoon to accelerate pesticide degradation.
   After four days of aeration, evaporation and solar oxidation, the
following results were obtained:
Pesticide
Atrazine
Dyanap
Treflan
Ami ben
Paraquat
Concentration
Initially After Treatment
40
24.034
1629
894
3.1
7.4mg/l
5 M8/1
5 eg/I
15.1 Mg/1
0.260(4/1
Removal
Efficiency, ^
81.5
99.9 +
99.7
98.3
91.6
Powdered activated carbon was injected into the lagoon at a self-
flocculating concentration of  1,000 mg/1  (Table 3). This carbon
dosage produced a clarified effluent with a suspended solids con-
centration of less than 10 mg/1. Consequently, post-filtration was
not required prior to discharge. A carbon contact period of 4 hr
was employed following a 48 hr clarification period.
  Powdered carbon  addition further reduced pesticide concentra-
tions in the  lagoon as shown below.
Pesticide
Atrazine
Dyanap
Treflan
Furidan
Concentration
After Aeration After Carbon
7.4
5
5
15
0.06 mg/1
0.5 /ig/l
O.I 7 Mg/1
.01 Mg/1
Carbon Removal
Efficiency, It
99.2

96.6
99.3 +
                                                           The treated water from the lagoon was then pumped behind a
                                                         sandbag  carbon/bentonite  impoundment  constructed  in  the
                                                         stream.  Powdered activated  carbon was injected into the im-
                                                         pounded waters to further reduce Treflan and Atrazine levels below
                                                         detection limits to 0.1  /tg/1 and 0.01 mg/1 respectively.
                                                           Static aquatic bioassays were run on 5 gal samples collected from
                                                         the lagoon and creek  impoundment following carbon  treatment.
                                                         Perch fmgerlings were used in the bioassay because they were in-
                                                         digenous to the area. Five fmgerlings were placed in each bioassay.
                                                         and deaths were recorded with time. A control bioassay was run
                                                         parallel to the test bioassays to identify potential interferences due
                                                         to oxygen deficiencies  or organism sensitivities and to isolate these
                                                         factors from toxicity effects.
                                                           The waters in the area (an old strip mining area) had a pH of 4 to
                                                         5; the pH had to be adjusted to  7  to avoid  a shock to the fith.
                                                         Elevating the pH formed an iron  hydroxide floe which may have
                                                         removed some additional toxic materials via chemical precipitation.
                                                         Another potential interference with these bioassays may have been
                                                         the introduction of air into the bioassay cells resulting in stripping
                                                         of volatile toxics.
                                                           Nevertheless, the bioassays indicated that no accutely toxic
                                                         materials remained in the treated effluent lagoon or stream waters.
                                                         In addition, the concentrations of chemicals in the water were
                                                         reduced below any reported TL,,, or LCso (concentration producing
                                                         50% organism  mortality). Survival times were recorded over 360 hr
                                                         and, prior to test termination, over 50% of the fish population was
                                                         surviving.
                                                           A  pool located in Shoal Creek, one mile downstream from the
                                                         lagoon discharge point, contained  perch fingerlings. This pool was
422
CASE HISTORIES

-------
                                                                 Table 1
                              Chemical Quantities, and Toxicological and Physicochemical Properties of Products
Pesticide
Trade Name
Treflan
Sutan
Surflan
Paraquat
N-Serve
Lorsban
Lasso
Eradicane
Furloe-C'hloro IPC
Furadan
Lorox
Lexone
Bbdex
Amiben
Dyanap
Aatrex
AUa-lox

B 1088
Banvel K
Bassagran
Randox
Total
Chemical Class
Dinilroaniline
Thiocarbamale
Dinitroaniline
Pyndiljuin
Pyridme
Org anophuspha le
Acolamllde
Thiocarbanule
Phenylcarbamate
Carbamate
Phenyurea
s-Tnazine
s-Triazmc
Bcnzoic acid
Phttialic acid
s-Triazme
Chlor. Hydrocarbon
& organophosphorus

Benzole acid
Benzothudiazm
Acetamide

Quantities
(gal)
530
140

56
75
310
70
40
10




190
370
1.355
60

4
55
200

3,465
(Ib)


80


1,100



1,100
2,000
1,500
200
4,000

850





250
11,080
Solubility
» 25°C TLV
Water (mg/l)

E
NS
S
NS
2
DISP
NS
NS
700
75
1,200
NS
S
S
70
40


SS
620,000
N

Solvent (mg/m )
Xylene

Ethanol
NS' 0.1
Xylene 10.0
Methanol 0.2
Monochlorobenzene 0.75

Xylene 100.0
N 0.1
Xylene
Ethanol
0.5
Alcohol
Ethanol
18,000 Methanol
S
Alcohol

Xylene
Ethanol


LD,0 (m
Oral
(rat)
3,700
3,690-4,500
10,000
150
2,140
2,000
1,800
2,000
3,800
11
1.500-4,000
4,000
334
3,400-5,620
232
3.080
2,000


2,500
1,480
750

l/kl)
Dermal
(rabbit)

4,640


5.000
4,000

3,830
10.200
10,200

2,000
7.200
3,160
400
10,200
8,000


2.000
2,500


TLM
96 -hi
(mg/l)
0.02


10.0




10.0
0.11




0.10
1.0







                               'DISP - disperiible; N • negligible; NS - not soluble; S - soluble; SS - slightly soluble; E • forms emuliion.
                                                                 Table 2
                                   Initial Contaminant Levels by Location and Decontamination Yardsticks
                                       TOC                                      Atrazine
                Sample Location & Date pH    (mg/l)    Alkalinity    Oil A. Grease Suspended Solids     (mg/l)
                  Dyanap
                  (W5/1)
Treflan
 (Hi/I)
Amiben   Paiaquat
 Mn     (mg/l)
               Burn Site Collection     7.1
                Pit No. 2, 4/29/80

               Lagoon, 4/29/80
                (Note: 4-days treated)  4.4
                                       400
                                                160
                                                           108
                                                                                 40
                                                                                         24,034     1,629      894     3.1
           7.4      <5
                                                                                                            15.1     0.260
Creek, Station 4,
4/29/80
(Note: 4-days treated) 7.1 22 150
Decontamination
Yardsticks 4.0* 10 0-100


5 34 4.9 121.7

5 20 <0.1 <100


7.5

<100


16.9

<100


0.340

<0.1
               *Low pH due to strip mine waters in irea.
                             Table 3
 Treatability Alternatives Resulting in Selection of Powdered Activated
            Carbon Addition for Water Treatment
                             Atrazine  Treflan Contact Time  Sludge Volume
     Treatment Method   Dosage Level   (mg/l)   OJg/D     ("')      (% volume)
Untreated Lagoon Sample
PAC"
PAC
PAC
PAC
H,0,b
HjO,
11,0,
N.,COjC

7.4 5.0
24, 500 mg/l <0.1 <0.1
l.OOOmg/1 <0.1 <0.1
5,000 mg/l <0.1 <0.1
15,000 mg/l <0.1 <0.l
0.03% <0.1 0.18
0.3 % <0.1 <0.1
3.0 % 0.11 <0.1
5 mVl 0.26 <0.1
saturated solution

10.5
0.5
0.5
0.5
15.0
15.0
15.0
24.0


13
0.5
3
8
4
4
4
20

 Aqui Nuchar powdered activated carbon, particle size 2 microns.

 Hydrogen peroxide-rejected because of potential formation of oxidized toxic by-producti and high
 sludge volume.

CSoda ash-rejected because of inadequate treatment and high sludge volume.
utilized for an in situ bioassay. No adverse effects were monitored
during cleanup operations. Treated water was released after con-
firming  tests  indicated that  concentrations  of monitored con-
taminants had been reduced below maximum allowable concentra-
tions (Fig. 5 and Table 4).
  The 2300  gal of  concentrated carbon/bentonite  slurry were
pumped from the lagoon bottom into expoxy-lined 17H drums for
disposal. Unpumpable residues (500 gal) were treated with soda ash
for alkaline hydrolysis  of the  remaining pesticides.  Activated car-
bon was blended into the remaining  lagoon bottom sludge, and a
24-mil lagoon  liner  was folded  in  half to  entomb the  treated
residues.

Contaminated and Uncontaminated
Building Materials

  Personnel wearing full protective  clothing carefully segregated
uncontaminated  building materials  from  contaminated building
materials. These materials were staged where conventional  high
lifts and open bed trucks could be used to load and transport them
away from the contaminated area. Consequently, only 29 yd3 of a
total of 131  yd3 of  contaminated and uncontaminated building
materials  were  disposed of  as  hazardous  wastes. Special  ar-
                                                                                                            CASE HISTORIES
                                                                423

-------
                                                            Figure 4
                                 Concentrated Pesticide Runoff Collected in Interceptor Pits it Burn Site.
                                     Residues from pit I transferred to pit 2. Floating pump used to
                                         transfer runoff into 17H drums by personnel wearing
                                                      full protective clothing.
rangements  were made with the Illinois EPA to transport these
materials  in bulk  containers  in 24-mil polyethylene capsules,
thereby minimizing packaging, transport and disposal costs.

Concentrated Leachates and Soil Matrices at Burn Site

Removal
  A total of 26,675 gal of highly concentrated pesticide leachates
and soil were transferred into epoxy-lined drums (Figs. 6 and 7). At
this time, regulations allowed landflUing of liquids in permitted
hazardous waste landfills. This method of disposal was selected to
minimize disposal costs at the request of the insurance company.
  At the time of this writing, the landfill to which these toxic liquid
wastes were sent is now experiencing a  leachate problem and will
probably  become  the  target of a future remedial  action  site
cleanup. In retrospect,  it would  have been more cost-effective in
the long run to have incinerated these wastes.
                                                        In Situ Detoxification
                                                          A total of 40,000 ft* of soils containing traces of agricuitunl
                                                        chemicals up to depths of 3 ft were detoxified in place. Soda aril
                                                        and powdered activated carbon were disced and  plowed into the
                                                        soil. Soda ash was applied periodically to maintain a soil pH near 9.
                                                        A water mist was applied  to activate the soda ash in soil.
                                                          Several physical/chemical processes were activated by this treat-
                                                        ment:  some pesticides  were degraded   by soda  ash alkaline
                                                        hydrolysis; the powdered activated carbon mitigated odors and ab-
                                                        sorbed agricultural chemicals, thereby preventing additional con-
                                                        taminant migration; the black carbon absorbed solar energy, thai
                                                        elevating  soil temperatures and  catalyzing pesticide destruction;
                                                        and periodic discing  and aeration of soils accelerated further
                                                        pesticide degradation by solar oxidation and evaporation. The ete
                                                        was monitored until the degradation of pesticides decreased beta*
                                                        0.1 mg/1.
                                                             T«bk4
                Final Pettldde Concentration! and Fish Bloanayt Uicd to Confirm DecoafaniMtton tx
                                                                                             iVarttottcks
                              Simple Detcription
                                                 Alrulne Ticflin Dyinip Furldin       Number of   Pitch Ftitftriini SurrtnJ Ttent
                                                       04|/l!  (M/l)  (m|/ll  pll  Flr>|tibnfl Tilled        (rut
Control
Creek prior 1o treatment
Treated lagoon walerl
Treated creek impoundment No.

«.»
006
1 water!  M
I.I
>JJ6
>J)6
 424
CASE HISTORIES

-------
                            Figure 5
   Static Bioassay Tests Run on Treated Lagoons and Stream Samples
 Resulted in Perch Fingerlings Surviving the 96-hour TLm Concentrations.
Contaminated Product Residues

  A total of 500 gal of product were recovered in their original con-
tainers. Due  to obvious potential third-party liability problems,
these materials were not saleable and were disposed of at a hazar-
dous waste disposal facility.
CLEANUP COSTS VERSUS
LIABILITY TRADE-OFFS

  The degree of decontamination accomplished should have reduc-
ed potential life safety threats and environmental impacts to accep-
table limits. In the experience of the author, hazardous material
cleanup costs range from 1 to 10% of potential personal injury suits
and/or property damage claims. Other elements entering into the
decision criteria used to determine the extent of cleanup included:
available financial resources; government regulations; local media
and public  hazard perception; local political/regulatory inter-
pretation of laws; and background contamination considerations.
  Because of the potential multi-million-dollar liabilities associated
with this incident, pesticide levels in  soils and waters were decon-
taminated below  the acceptable environmental health limits and
below background concentrations typical of the watershed  (Table
4).
  Resources committed to this  project involved around-the-clock
emergency operations for 21 days and 3,200 man-hours of effort by
60 engineers, scientists and engineering technicians. Losses incurred
by  the owners (including chemicals, buildings and this cleanup)
totalled nearly $500,000.  Fortunately, all losses were insured, and
the owners were reimbursed. In  addition, because the site was ade-
quately decontaminated, no legal actions have been filed.

CONCLUSIONS
  Liabilities,  cleanup costs and  life safety hazards can be reduced
by using experienced personnel with the ability to rapidly evaluate
risks at hand and to select and implement engineered isolation, con-
tainment, recovery, treatment and disposal procedures. Quickly in-
tegrating the engineering investigation with an engineered solution
provides a cost-effective solution to hazardous material problems.
  The success of this project included: no injuries or significant im-
pact to public health and welfare;  no  significant impacts to the en-
vironment;  and decontamination  of over 250,000  gal  of con-
taminated  surface waters  using powdered activated carbon
resulting in cost savings of over $180,000 when compared  to the
cost of using a mobile carbon treatment unit.
  Utilization of natural in situ pesticide destruction processes such
as photolysis  and  evaporation provided a cost-effective means of
detoxification of pesticide-contaminated water and soils. Alkaline
hydrolysis accelerated pesticide degradation in the soils and  water.
                            Figure 6
   REACT Personnel Packaging Dyanap, Treflan and White Atrazine
    Residues. Packaging Operations of Atrazine and Treflan Residues
      Using Polyethylene Bags Placed in 17H Epoxy-Lined Drums.
Fish bioassay screening of contaminated water from agricultural
fire runoff afforded a pragmatic method of determining the toxici-
ty reduction of unidentified oxidized chemicals below aquatic tox-
icity limits.

ACKNOWLEDGEMENTS

  The success of this project would not have been possible without
the exemplary actions of Richard and Roy Hewitt, Millers Mutual
Insurance, the Hillsboro Volunteer Fire Department, the citizens of
Hillsboro and Schram  City  and  Bill Busch  of the Illinois  En-
vironmental Protection Agency.
                                                                                                    CASE HISTORIES
                                                           425

-------
                             Figure 7
   Contaminated Residues Packaged in 17H Drums. 485 Drums Were
  Loaded Using Front-end Loaders and Transported by Covered Trucks
                     for Ultimate Disposition.
BIBLIOGRAPHY
  1. Sever,  W., el al.,  Illinois Pesticide Applicator Study  Guide, Uni-
    versity of Illinois, Urbana Champaign,  IL, 1975.
 2. Busch,  W.H.  and Renkes,  J., "Organic Chemical  Fire in  Illinois:
    Emergency  Response  and  Cleanup,"  Civil  Engineering-ASCE.
    Sept.. 1982, 62-65.
 3. Department of Health, Education and Welfare, "Occupational Ex-
    posure to Pesticides," N/OSH Criteria for a Recommended Standard,
    Cincinnati, OH,  1979.
 4. Encyclopedia of Occupational Health and Safety. McGraw-Hill Book
    Company. New York. NY, 1972.
 5. Hawley, G.G.. The  Condensed Chemical Dictionary.  9th ed., Van
    Nostrand Reinhold Company, New York. NY. 1979.
                                                            6. Hayes. W.J., M.D., Ph.D., Toxicology of Pesticides. The William
                                                               and Wilkins Company, Baltimore, MD, 1975.
                                                            7. Lindak, J.E.  and Haas, T.J., "Pesticide...11  Can Be a Problem u
                                                               Sea," Proc.  of 1980 National Conference on  Control of Hazardoui
                                                               Material Spills. Louisville, KY, May, 1980, 30-34.
                                                            8. Manufacturers Material  Data Safely  Sheets listed by product and
                                                               company;  Ciba-Geigy, Alpha-tox; Ortho Chevron Chemical Com-
                                                               pany, Paraquat; Dow Chemical Company, N-Serve; PPG Industries,
                                                               Furloe Chloro IPC;  Stauffer Chemicals, Eradicane;  Amchem Prod-
                                                               ucts,  Inc.,  Amiben;  Dow  Chemical  Company,  Lorsban;  Uni-
                                                               Royal. Dyanap; Ciba-Geigy, Alrazine (AATREX); BASF Wayndotte
                                                               Corporation;   Basagran;   Monsamo-Randox;  Monsanto,  Lasso;
                                                               DuPont, Lorox; DuPont, Lexone; Shell Oil Company. Bladex.
                                                            9  Marsh, J.R.  and  Phung, H T., "Disposal of Dilute Pesticide Solu-
                                                               tions." Proc. of 1980 National Conference on Control of Hazardous
                                                               Material Spills. Louisville, KY. May, 1980. 403-410.

                                                            10. Meister.  R.T.,  Farm  Chemicals Handbook.  Meister  Publishing
                                                               Company, Willoughby, OH, 1980.
                                                            .1. Pally.  F  A.. Ed., Toxicology. Industrial  Hygiene and Toxicology,
                                                               Volume II. 2nd ed , Interscience Publishers, New York, NY,  l%3.
                                                            12. "REACT, Computer  Assist  Program," REACT Corporate Response
                                                               Center, St. Louis, MO,  1980.
                                                            13. Ryckman,  D.W. and Miller, David B., "Case Histories of Emergency
                                                               Action Report) from Men on the  Firing Line," Hazardous Materials
                                                               Spills Seminar, AWWA  Conference. Anaheim, CA. 1977.
                                                            14. Ryckman,  D.W.,  Ryckman. M.D.  and Miller,  D.B., "Emergency
                                                               Action Response  for Hazardous Substances,"  American Defense
                                                               Preparedness  Association Energy  Environment  Conference, Kansas
                                                               City. MO, 1977.
                                                            15. Ryckman,  D.W. and Ryckman, M.D., "How to Cope with Hazardous
                                                               Material  Spills  that Threaten  Water  Supplies,"  1979  AWWA
                                                               Annual Conference and  Exposition,  San Francisco, CA, 1979.
                                                            16. Ryckman,  M.D., "Competency  Factors  in  Hazardous  Materials
                                                               Spills Response." Fourth Inland  Spills/Hazardous Waste Disposal
                                                               Conference, Cleveland, OH. 1979.
                                                            17. Ryckman,  M.D.,  Wiese. G.T and Ryckman,  D.W.,  "REACTs
                                                               Response to  Hazardous  Material Spills." Prof.  1978 National Con-
                                                               ference on Control of Hazardous Material Spills.  Miami, FL, 24-26.
                                                            18. Ryckman,  M.D.,  Rains, B A. and  Miller,  R.L., "Flammable Liq-
                                                               uid Spills—Response and Control,"  Proc. of I960 National Confer-
                                                               ence of Hazardous Material Spills. Louisville, KY. May, 1980, 14-22.
                                                            19. Ryckman,  M.D.,  el  al..  "Toxic Corridor Projection  Models for
                                                               Emergency  Response."  presented at  the  Transportation  Research
                                                               Board 61st Annual Meeting. Washington. DC. Jan., 1982.
                                                            20. Ryckman,  M.D., et al., "Emergency Response to a  Major Agricul-
                                                               tural  Chemical Warehouse  Fire,"   Proc.  of the 36th  Industrial
                                                               Waste Conference. Purdue University, Lafayette,  IN,  May 1981, 212.
                                                            21. Standard Methods for the Extermination  of  Water  A Wastewater.
                                                               14th  ed..  Part  800—Bioassay  Methods  for  Aquatic  Organism.
                                                               American Public  Health Association. Washington, DC, 1975.
                                                            22. Stroud, F.B., et al.,  "Kenco Chemical and Manufacturing Corpora-
                                                               tion's  Pesticide Fire, Jacksonville,  Florida," USEPA.  Region  IV,
                                                               Atlanta. GA.  1979.
                                                            23. Wilkinson, R.E., et  al..  State-of-the-Art Report. Pesticide Disposal
                                                               Research. USEPA. Sept., 1978.
426
CASE HISTORIES

-------
      SUPERFUND  REMOVALS IN  REMOTE AREAS  OF THE
WORLD: PACIFIC  ISLAND IMMEDIATE REMOVAL PROJECT

                                           MICHAEL H. RIDOSH
                                              Roy F. Weston, Inc.
                                               in Association with
                                                 Tetra Tech, Inc.
                                           San Francisco,  California
                                             WILLIAM E. LEWIS
                                            CHRISTOPHER VAIS
                                         Emergency Response Section
                                   U.S. Environmental Protection Agency
                                           San Francisco,  California
                                           LT. JACK A. KEMERER
                                   U.S. Coast Guard, Pacific Strike Team
                                               Novato,  California
 INTRODUCTION
   Region 9 of the USEPA recently completed a Superfund immed-
 iate removal action at 39 hazardous waste sites in the Territory of
 Guam, the Commonwealth of  the Northern Mariana Islands
 (CNMI),  and in the  Trust Territories  of the Pacific Islands
 (TTPI). The  logistics associated with a project of this magnitude
 were complex and required 10 months of careful planning to over-
 come numerous obstacles. All site work was completed in 10 weeks
 at an estimated cost of $ 1.4 million.
   Guam is an unincorporated territory of the United States under
 the jurisdiction of the Department of the Interior. It is  also the
 southernmost and largest of the Mariana Islands, situated about
 2,170 km south of Tokyo and 5,000 km west of Honolulu (Fig. 1).
 The island covers an area of 541 km2 and has a total population of
 105,816, including approximately 22,000 military personnel and de-
 pendents. The economy is predominantly agricultural, with a high
 production of fruit, vegetables, eggs, pork, beef and poultry. Fish-
 ing and tourism also constitute  a large portion of Guam's econ-
 omy.
   The Commonwealth of the Northern Mariana Islands (CNMI),
 excluding Guam, is composed of 16 islands in the Western Pacific,
 including the three largest: Saipan, Tinian  and Rota. The chief
 settlement, the administrative center for the commonwealth gov-
 ernment and  the government of the Trust Territories of the Pacific
 Islands (TTPI) are in Saipan. Saipan, the largest of these islands,
 has an area of 122 km2 and a population of 14,885. Tinian, the
 second largest island, covers 83 km2 and has a population of 899,
 while Rota covers 83 km2 and has a population of 1,274. The major
 industry is the production of vegetables, beef and pork. Tourism
 also contributes to the economy of the commonwealth.
   The Trust Territory of the Pacific Islands consists of the Caroline
 Islands and  the Marshall Islands in the Western Pacific and the
 Mariana Islands in the Northern Pacific. The 2,185 islands, 84 of
 which are inhabited, are grouped into seven  administrative dis-
 tricts (Fig. 2). The territory lies within an area known as Micro-
 nesia. The Marianas District achieved separate status as the Com-
 monwealth of the Northern Mariana Islands and remains  legally a
 part of the trusteeship until the Trust Territory is dissolved. The
 Marshall Islands District drafted its own constitution in 1979. Of
 the five districts of the Caroline Island group, the four districts of
 Yap, Truk, Ponape and Kosrae ratified a new constitution to be-
 come the Federated States of Micronesia in May, 1979.
  In the seventh district, the Palau District, a referendum approved
a proposed local constitution in July, 1979. It came into effect on
Jan. 1, 1981, when the district became the Republic of Palau. Elec-
tions are currently being held to establish independence for the
Marshall Islands, the Federated States of Micronesia (Yap, Truk,
Ponape, Kosrae) and the Republic of Palau.
  The total land area  of TTPI  is 1,300 km2; the largest islands
are Babelthuap (367 km2) in the Palau District and Ponape Is-
land (330 km2) in the Ponape District. The total population in 1980
was 116,974,  distributed as follows: Marshall  Islands:  31,041;
Palau:  12,177; Ponape: 22,319; Truk: 31,742; Yap: 8,172; and
Kosrae: 5,522. The area of the TTPI superimposed over a map of
the mainland is shown in Figure 3.
  The chief crops are coconuts,  breadfruit, bananas, taro, yams,
cocoa, pepper and some citrus. Subsistence crop production pre-
dominates and, except for copra, little is marketed.
BACKGROUND

  USEPA Region 9 has had an environmental program for many
years aimed at protecting the fragile environment of the Pacific
Islands. The Agency gradually became aware of the uncontrolled
hazardous waste sites on Guam, in the CNMI, and in Micronesia
through its contacts with local government officials. Many of these
sites were the subject of numerous reports and studies dating back
to 1978.
  USEPA contractors  have surveyed hazardous waste sites in the
Pacific Islands at least three times since 1978.  These surveys docu-
mented numerous localized acute problems involving pesticides and
PCBs, but little progress was made toward solving these problems,
due to the limited  resources and technical expertise of the  local
agencies.
  In 1981, when CERCLA was implemented, the island govern-
ments hoped to obtain  cleanup of all sites  under the remedial pro-
visions of the  new law. The rules covering selection of individual
states' highest priority sites required narrowing the scope of the re-
medial program to a few of the highest priority sites for inclusion
on the National Priority List (NPL). Of the 32 sites originally con-
sidered for the NPL, only  11 were included on the list. The re-
maining sites did not score sufficiently high  for inclusion.  There

NOTE: This paper represents work that was carried out in part with USEPA funding under the
USEPA Contract No. 68-01-0669. The opinions expressed in the paper are those of the authors
and do not necessarily reflect official policy of either the USEPA or USCG.
                                                                                          CASE HISTORIES
                                                     427

-------
               J-'TRUSTTERBiTCmirOFTHS PACIFIC ISLANDS
                                             ENVIRONMENTAL PROTECTION AGENCY REGION IX
                                                           Figure I
                                          U.S. Environmental Protection Agency Region IX
remained, therefore, a continuing concern about the public health
risk associated with the non-NPL sites.
  Recognizing this, in November, 1982, Region Nine's Emergency
Response Section, at the request of the Region's Office of Terri-
torial Programs, began an  evaluation of the  non-NPL sites for
possible planned removal action. All previous studies wered re-
viewed,  additional inquiries were made by mail and, in March,
1983, a  member of the Region Nine Technical Assistance Team
(TAT) conducted a survey of these sites. At the same time, the NPL
sites were surveyed for changed conditions.
  There were 32 known or suspected sites on the islands of Guam,
Saipan, Koror in the Republic of Palau, Moen in Truk State, Yap,
Kosrae,  Ponape and the islands of Jaluit, Eniwetok, Majuro and
Ebeye in the Republic of the Marshall Islands. An additional seven
sites were brought to the USEPA's attention upon arrival in the
islands. The majority of the sites in Saipan  and Micronesia  were
storage yards containing PCB transformers.
  There were also a number of pesticide storage houses filled with
deteriorating containers of banned or restricted use pesticides and
some small dumps of industrial chemicals. The sites on Guam in-
cluded a transformer storage yard, a hospital chemical storeroom,
pesticide storage sites, a storage yard of deteriorated acid drums
and storage sites of industrial chemicals.
  Only a few of the sites appeared  to present difficult technical
problems. The  staff anticipated, and were prepared to  live and
work under, adverse conditions.  The USEPA's primary concern
for the success of this removal action was logistics. In order to in-
sure success, the USEPA had to purchase and ship all equipment
and supplies overseas in advance of departure. Once on an island,
                                                      the cleanup crew had to package the waste, containerize it for ship-
                                                      ment and make the necessary shipping arrangements for transpor-
                                                      tation to and disposal at an appropriate waste site. The problem
                                                      was exacerbated by poor site access, lack of security at the sites and
                                                      waterfront, substandard  roads, erratic weather conditions  which
                                                      lead to erratic shipping schedules and the shipping requirements for
                                                      incompatible wastes.
                                                      PRELIMINARY ASSESSMENT
                                                         The decision to  reevaluate all  of the Pacific Island sites was
                                                      made late in 1982. It was immediately apparent that the available
                                                      information was not adequate to  evaluate current site conditions.
                                                      In March,  1983, a member of the Region IX TAT conducted a Pre-
                                                      liminary Assessment  (PA)  of all  of the reported sites except for
                                                      those on Kosrae. TAT member Erwin Koehler spent one month
                                                      visiting the sites, investigating local resources and logistical sup-
                                                      port capabilities and researching potential health effects associated
                                                      with  these sites.  The PA  resulted in  the  following information
                                                      which was used in planning the project:
                                                      •Detailed inventory of the kind and amount of waste at each site
                                                      •Analysis of the cleanup needs at each site
                                                      •Analysis of the environmental and health threats at each site
                                                      •Estimate of the number of empty drums required
                                                      •Estimate of the number of work days required on each island
                                                      •Evaluation of the logistics  required on each island  to ship out
                                                        supplies and equipment and return drummed waste
                                                      •Evaluation of the site safety and personal protection requirements
                                                         With the PA complete, the project moved into the planning
                                                      stage.
428
CASE HISTORIES

-------
                                             •:iŁ? MARIANA
                                               '    ISLANDS
                                   MARSHALL
                                   ISLANDS
                        Palau
                           District
\ Yap
 \District
                                    CAROLINE      i S'L 'AN D s
Truk
 District
Ponape
                                         ENVIRONMENTAL PROTECTION AGENCY  REGION IX

                                                         Figure 2
                                              Trust Territory of the Pacific Islands
 PLANNING
  The planning for this project involved a number of tasks; prin-
 cipal among these were the need to:
 •Reevaluate all sites and determine the number to be dealt with as
 immediate removals
 •Prepare a funding request
 •Develop a removal contract
 •Resolve logistics issues
 •Coordinate with local governments
 •Assign personnel
  The reevaluation of the 33 sites investigated during the prelim-
 inary assessment was completed in April,  1983. The TAT survey
 found that, in general, 32 sites qualified for planned removal
 action. Furthermore, several sites, including some NPL sites where
 conditions had deteriorated,  warranted  immediate removal. This
 information, along with the recent apparent policy change to treat
 planned removals as immediate removals and with the exemption
 from matching fund requirements enjoyed  by the territories under
 the Omnibus Territories Act, led the USEPA to propose these ac-
 tions as immediate rather than planned removals (even though
 some sites  represented potential,  rather than existing threats). It
 should also be  noted that emergency response operations cannot be
 considered routine when conducted in remote areas. Public health
 and welfare would  be best protected by removing as many poten-
 tial emergency situations as possible in one sweep  through the is-
 lands, minimizing the need for future responses.
  A funding request was prepared and  submitted in May, 1983.
 The initial request was for $795,000. However, at the conclusion of
 the planning process, the USEPA realized that it had underesti-
 mated the amount of hazardous waste to be removed and, as a re-
sult, the cost  of the project. In  December, 1983, an additional
$602,745 was requested, thus raising the estimated project cost to
$1,397,745 (Table 1).
                       The removal contract, developed by the USEPA headquarters
                     Procurement and Contracts Management Division (PCMD), was
                     unlike the usual emergency cleanup contract used by the Agency
                     at the time. The contractor, Unitek Environmental Services, Inc.
                     in Honolulu, HI, was hired to arrange the procurement and ship-
                     ment of supplies to the islands, the shipment of waste back to the
                     mainland and disposal. Instead of the usual time and materials
                     contract, PCMD developed an indefinite delivery/indefinite quan-
                     tity type contract with a minimum contract amount of $541,533.00
                     and a maximum of $1,246,445.00. The required supplies and serv-
                     ices were ordered through  the issuance of Delivery Orders by the
                     contracting officer or the OSC's.
                                              Figure 3
                        Area Comparison Between the U.S. and the Trust Territory of the
                                            Pacific Islands
                                                                                              CASE HISTORIES
                                                                                                                      429

-------
                           Table 1
                   Estimates of Removal Costs
                                                                                     Table 2
                                                           Partial List of Supplies Purchased and Shipped for Use on the Pacific
                                                                                  Islands Project
Item
                               Maximum Estimate
Personnel support:
  Government travel/per diem      $  63,300
  Government labor              $  88,000
  Contractor travel/per diem       $  15,000
  Contractor labor               $  42,432
Materials and shipping            $ 242,935
Transportation and disposal of waste $ 876,078
Contingency                     $  70,000
  Total                         $1,397,795
   The final contract cost would be a function of the amount and
 type of waste removed. Because the costs of the supplies and serv-
 ices had been agreed upon in the contract, the OSCs were freed of
 much of the paperwork associated with a time and materials con-
 tract.
   As with any project located in a remote area, logistics were the
 most critical element for a successful operation. The plan we devel-
 oped called for all of the supplies to be delivered to  the USCG
 Pacific Strike Team in Novato, CA.
   All of the supplies (Table 2) were stored in an aircraft hangar.
 The hangar was subdivided  into spaces corresponding  to each is-
 land, where the cleanup staff would be working. The supplies were
 then divided among all of the islands and placed in the appropriate
 spaces. The division of supplies was based upon a very careful eval-
 uation of all work to be done on each island.
   Items such as boots, suits,  gloves, respirator canisters, etc.,
 were packed into 55 gal drums which were then placed inside of 85
 gal overpack drums.  The idea  was to consolidate the  supplies as
 much as possible because shipping rates are calculated on the basis
 of weight or volume, whichever is most advantageous to the ship-
 per. Once the supplies were consolidated, they were packed into
 containers for shipment directly to each island.
   The staff considered and rejected the idea of staging supplies
 and equipment on Guam or Hawaii because shipments could be
 made directly to all the islands from the West Coast. Two steamship
 companies were used. Of the two carriers serving Guam, United
 States Lines was utilized because they were  the only carrier who
 would return the waste to the West Coast from Guam.
   Philippines Micronesia & Orient Lines (PM&O) provided direct
 services to  Saipan and Micronesia. PM&O  had no reservations
 about returning wastes to the West Coast. The supplies were con-
 signed to local government officials agreeing to provide secure stor-
 age pending the USEPA's arrival.
   Waste to be returned to the  mainland for disposal was packed
 in drums  and containerized for shipment back to the mainland.
 The amount of waste shipped  off each island  by hazard class is
 shown in Table 3.
   Only  one shipping related problem with the potential to disrupt
 the project was encountered. On the return voyage from Guam to
 the mainland, USL ships go  via Japan. Initially, the Japanese gov-
 ernment denied  permission to ship OCB-contaminated waste
 through Japanese territorial waters. However,  the Japanese gov-
 ernment relented when they were advised that  no containers  of
 PCB wastes would leave the ship for any reason. Had they not re-
 lented,  the  USEPA would have had to forward the Guam PCB
 wastes to  Saipan and ship them back to the  mainland  from there
 via PM&O Lines.
   All of the supplies shipped to the Islands arrived in good shape.
 All of the waste returned to the mainland arrived without inci-
 dent except for one container in which a small quantity of  PCB-
 contaminated diesel rinsate leaked.  This container was decontam-
 inated and returned to service.
                                                                   Hem Description
                                                                                                                         Quantity
                                                         Digital Sphygnomanometer and Indicator
                                                         Electronic Thermometers w/covers
                                                         Centec Specific Ion Probe
                                                         Plastic drum liners
                                                         Halazone (Bottles) REI
                                                         Sodium Phosphate (I Ib. ea.)
                                                         Sodium Phosphate (500 gm ea.)
                                                         Mosquito Repellant (aerosol)
                                                         Trimelhylpentane (4 per case)
                                                         Longhandled Decon Brush
                                                         Plastic Basin
                                                         Visquene, 6 mil x 20 ft x 100 ft
                                                         Spray Paint, (red) aerosol
                                                         Spray Paint, (yellow) aerosol
                                                         Spray Paint, (orange) aerosol
                                                         Powdered drink mix (salt replacement type)
                                                         55 gal drum gaskets
                                                         5/8 in. Bolts for Locking Rings
                                                         Flat Shovels
                                                         Pointed Shovels
                                                         Bug Spray
                                                         Sijal 1896 one piece green suit
                                                         Charkate #6001B Tyvek Coverall
                                                         Edmont *34-300 Inner Gloves
                                                         Pioneer /TAK-22 Nitrile Gloves
                                                         Fab Ohio 0320-840 Outer Booties
                                                         Ranger Neoprene Steel Toe Boot
                                                         Duct tape, 2 in.
                                                         Keystone W493TII Drum Truck
                                                         McMaster-Carr 02696T1 Drum Handler
                                                                      *3401T2 Drum Sling
                                                                      *4239K3 Hand Pump
                                                                      #4239K4 Repair Kit
                                                                      W756A3 Drum Cutter
                                                                      W756A4 Extra Blades
                                                         Hudson tnS2TTl Sprayer (3 gal)
                                                         Emergency Medical Technician Kit
                                                         Mechanics Tool Kit
                                                         MSA 0457095 Respirator (Gas Mask)
                                                         MSA #GMC-SS-I Canister 477713
                                                         MSA #448966 Pesticide Canister
                                                         Vermiculite 14 Ib bag, Medium Grain
                                                         Overpack Drum, 85 gal
                                                         Overpack Drum, Liner
                                                         Reagents for PCB Analysis (Centec)
                                                         Comealongd ton)
                                                         Drummaster #V-1-F Forklift Drum Grabber
                                                         Labels
                                                         Placards
                                                         Manifests
                                                         New 17E Pails (5 gal) Bung
                                                         Picks
    2
    2
    4
  990
   24
   26
   24
   12
    2
   50
   50
   12
   35
   35
   35
   40
   40
   10
   4
   2
   12
  306
  965
   72
  965
  965
   46
  250
   2
   2
   2
   2
   4
   2
   4
   2
   2
   2
   18
  389
   60
  150
  160
  160
1,100
   2
   2
1,600
  150
  too
   20
   2
                                                           Each team had a pair of equipment boxes with tools instruments
                                                         and personnel gear that was shipped from island to island by air
                                                         cargo. While expensive, air cargo costs were much less than the cost
                                                         of multiple sets for all of the equipment required.
                                                         Local Government

                                                           From the beginning, the USEPA intended to encourage partici-
                                                         pation  by local governments to the extent of their  capabilities.
                                                         Each was asked to support the removal actions by providing:
                                                         •Secure storage of supplies prior to team arrival
                                                         •Certain items of heavy equipment
                                                         •Unlimited access to sites to ensure meeting time schedules
                                                         •Transportation for the team is not commercially available
                                                         •Disposal sites for non-hazardous waste
                                                         •Community relations
430
CASE HISTORIES

-------
                                                            Table 3
                                          Summary of Hazardous Waste Shipped for Disposal
                                   (Does not include material which was treated or disposed of locally.)
YAP
PCB liquid
Waste DDT
PCB contaminated debris
PCB capacitors
Poison B waste, lab pack
Waste carbamate pesticide, lab pack
Hazardous waste NOS, lab pack
PCB transformers

PONAPE AND EBEYE
PCB contaminated liquid
PCB liquid
Waste, poison B, lab pack
PCB contaminated debris
Corrosive wastes, lab pack
Waste ammonium hydroxide
Empty PCB drum
Hazardous Waste NOS, lab pack
Waste asbestos
PCB transformers

KOSRAE
Contaminated debris
PCB contaminated liquid .
Waste, poison B, lab pack


TRUK
PCB contaminated liquid
PCB liquid
PCB contaminated debris
Empty PCB drums
             SAIPAN
 5 drums      Waste glacial acetic acid
 8 drums      PCB liquid
  1 drum      Waste sodium arsenite
  1780 Ib      PCB contaminated debris
 5 drums      PCB contaminated liquid
21 drums      Waste, poison B, lab pack
 2 drums      Waste methyl alcohol
  7320 Ib      Corrosive liquid waste, lab pack
             PCB transformers
             GUAM
10 drums      PCB contaminated liquid
 4 drums      pCB liquid
 2 drums      PCS contaminated debris
  1 drum      PCB transformers
  1 drum      PCB capacitors
 5 drums      Solid hazardous waste, NOS
  1 drum      Solid waste poison B
  1 drum      Liquid hazardous waste, NOS
 4 drums      Waste flammable liquid, lab pack
  2320 Ib      Waste sodium silicofluoride
             Waste corrosive liquid, poison B
             Waste hydrofluorosilicic acid
             Waste nitric acid, lab pack
 3 drums      Waste corrosive liquid, NOS
  20 pails      Waste calcium hypochlorite
 5 drums      Waste oxidizer solid, NOS
             Waste oxidizer, poison B
             Waste formaldehyde solution
             Waste, poison B, lab pack
12 drums      Waste corrosive liquid, lab pack
22 drums      Waste oxidizer, lab pack
 2drums      ,.. ĄĄTDr.
 7 drums      MAJURO
  1 drum
65 drums
 13,550 Ib
 2 drums
 2 drums
 6 drums
  1 drum
  1 drum
60,730Ib
 7 drums
14 drums
 3 drums
 7,550 Ib
 3,7801b
 9 drums
 7 drums
 4 drums
 4 drums
 3 drums
 2 drums
17 drums
  1 drum
 7 drums
 3 drums
 2 drums
  1 drum
 4 drums
27 drums
 2 drums
  1 drum
PCB transformers


PALAU
PCB contaminated liquid
PCB liquid
Flammable liquid waste, lab pack
PCB transformers
19,290 Ib



2 drums
15 drums
1 drum
8680 Ib
Waste DDT
PCB liquid
PCB contaminated solids
PCB contaminated liquid
Waste, poison B, lab pack
Waste solid, poison B
PCB transformers
PCB capacitators
9 drums
17 drums
121 drums
5 drums
2 drums
17 drums
10,300 Ib
150 Ib
  In the preliminary assessment, the USEPA had reviewed local
government support capability. During the planning process, each
island was contacted with a list of needs and a request to advise
the USEPA if they could commit the requested  resources or, if
not, to advise the  agency as to what they could provide. Two
months prior to departure, all local governments were recontacted
to determine if their commitments were firm. As a result of this
effort, the agency received the anticipated  level of support and, in
a few cases, obtained more assistance than expected.

Personnel

  All of the on-site work was performed by two teams made up of
the USEPA, USCG and civilian contractor personnel. Each team
consisted of one USEPA On-Scene Coordinator  (OSC),  three
members of the USCG Pacific Strike Team (PST),  one member
of the USEPA's TAT and one employee of Unitek,  the logistical
support contractor.
  Each team was designed to be self-sufficient. All of the team
members were qualified to work in level B  protection, and all were
familiar with the monitoring instruments, particularly the PCB test
kit. Finally,  all  were  experienced response  personnel and were
familiar  with hazard categorization, manifesting, site safety and
decontamination procedures. Moreover, they all had worked  with
PCBs, pesticides and industrial chemicals in the past.
              In addition to his usual duties, each OSC acted as the team leader
            and primary U.S. Government representative to the local govern-
            ment officials;  he also worked on-site with the other team mem-
            bers. The PST supplied each team with one member who acted as
            the emergency medical technician (EMT) in addition to his usual
            duties  as heavy  equipment  operator,  diesel mechanic  and/or
            trained responder. The TAT's primary function was to support the
            OSC in matters of safety and administration along with  on-site
            work. Individual TAT members' expertise in chemistry was  used in
            Guam several times. Unitek personnel were to handle all logistics,
            arrangements in the field and on-site work.
              On Guam, the sites were split into two groups, and the two teams
            worked independently of each other. When the work on Guam was
            complete, the teams separated: Team A went to the Marshall Is-
            lands, Ponape and Kosrae; Team B went to Truk, Yap,  Palau and
            Saipan.

            SITE SAFETY

              Personnel protection and safety was of primary concern on the
            project. The conditions under which the teams expected to be oper-
            ating offered a greater opportunity for injury or exposure to toxic
            materials than any removal any member had done recently.
              The risk of physical injury was enhanced because of the lack of
            adequate equipment  and greater reliance on manpower to move
            drums and transformers. The lack of purified drinking water, the
                                                                                                   CASE HISTORIES
                                                                                                                           431

-------
lack of water for bathing and limited sanitary facilities all had the
potential to result in debilitating illness.
  Perhaps the most serious health threat encountered on a daily
basis was heat stress. Despite the efforts of the EMTs to monitor
team members, everyone suffered the effects of heat stress on more
than one occasion, and one team member was incapacitated due to
dehydration. The team attempted to minimize  the potential for
heat stress by working at night and carefully monitoring each in-
dividual before, during and after work.
  Prior to departing from the mainland, the TAT had prepared a
site safety plan for each of the sites. The plans were reviewed upon
arrival in order  to incorporate changes in site  or working  con-
ditions. One such modification, introduced on Guam and used for
the duration of the project, was the modification of the decontam-
ination procedure to eliminate the use of washdown water as much
as  possible.  Washdown was  limited to  the SCBA  bottle  and
harness and to emergency situations involving  gross accidential
contamination. The so-called "dry decon" consisted  primarily of
discarding all external protective clothing after each site entry.
FIELD TESTING
  The Guam EPA had moderate  analytical laboratory capability.
Essentially,  there were no other laboratory  capabilities in the Trust
Territories.  Therefore, it  was necessary to do as much material
characterization in the field as possible to avoid long delays in de-
termining what handling, shipping and disposal techniques were re-
quired. There are basic hazardous categorization tests which are
done on any waste site (i.e.,  pH, flammability and general iden-
tification of physical properties). These tests alone would not  have
been sufficient for the Pacific Islands Project.
  Having team members with extensive chemical backgrounds was
very useful. This enabled  the team to sort  hundreds of bottles of
waste laboratory chemicals into the proper hazard  classes using
label identification.  The team was also able to identify an un-
marked drum of an unknown material (Freon) and to test a white
powder identified only as "Miracle" to determine whether it was a
pesticide or  a cleaning compound. This task was done with a small
amount of laboratory equipment and a knowledge of the physical
and chemical properties of the materials involved.
  In another case, volumetric analysis was used to determine the
strength of waste acids and eliminate concern that the  acid in open
topped  drums had stratified  (stratified acid presents a hazard if
heat is generated when the acid is mixed during recovery).
  The team's most useful  piece of field test equipment was a Cen-
tec PCB test kit which measures PCBs by causing the release of a
chloride ion which, in turn, is detected by a specific ion electrode.
Use of the test kit allowed the quantitative  categorization of trans-
formers, capacitors, transformer oil and soil into  one of three cate-
gories: non-PCB, PCB-contaminated and PCB (less than 50 ppm,
greater than 50 ppm but less  than 500 ppm and greater  than 500
ppm, respectively). The team was also able  to determine the extent
of PCB soil contamination.
  The kits performed very well, needing only replacement of bat-
teries and, in one case, the specific ion probe. They operated bet-
ter  in a cool hotel room than outside, where the  temperature was
85 °-87 °F and the humidity was close to 90  percent. The real prob-
lem involved reagents: hexane evaporated during shipment from
the mainland necessitated development of a calibration curve using
2,2,4 trimethylpentane as a substitute.
DISPOSAL OPTIONS

  The largest expense associated with the removal actions was the
shipment and disposal of hazardous wastes. There are no RCRA-
permitted hazardous waste sites anywhere in Micronesia, so al-
ternatives to disposal in a mainland hazardous waste site were very
limited. The options considered were:
•Incineration
•Neutralization
•Recycle/reuse
•Explosive destruction
•Landfill
                                                         Incineration was used to dispose of transformer oil with a PCB
                                                       content of less than 50 ppm. The oil was mixed in a 1 to 5 ratio
                                                       with diesel oil and used for fuel in the local power plants.  Over
                                                       200 drums of oil were disposed of in this manner. Shipment and
                                                       disposal on the mainland would have colst about $550 per drum,
                                                       so an estimated $110,000 was saved.
                                                         Neutralization of waste acid had been proposed at 2 sites. In one
                                                       case, the acid was  being used so there was no need to remove or
                                                       neutralize it. In the other case, the USEPA  proposed to neutral-
                                                       ize 900 gal of what was supposed to be rainwater contaminated by
                                                       hydrofluorosilicic acid. After testing the material, the team found it
                                                       to have a  pH of near 0 and  a normality of 4 to 6. Neutralization
                                                       was abandoned because the acid  was significantly stronger  than
                                                       expected, and the OSC felt that a neutralization reaction  could re-
                                                       sult in the formation of a vapor  cloud which could potentially
                                                       affect nearby residents.
                                                         Reuse or recycling of hazardous materials was only practical on
                                                       Guam. Here, the team  was able to recover two drums of caustic
                                                       soda and give them to a local chlorine bleach manufacturer. Many
                                                       of the "waste" laboratory  reagents  removed from  laboratory
                                                       chemical store rooms were found to be of value and were returned
                                                       to the clinical laboratory for use.
                                                         Some of the  waste sites yielded highly reactive materials  such
                                                       as picric acid, white phosphorous, metallic sodium, ether and/or
                                                       non-shippable containers of compressed gases. On Guam,  these
                                                       materials were turned over to a Navy Explosive Ordinance Dis-
                                                       posal Team who safely destroyed them at an  explosives range. On
                                                       Yap, a number of containers of picric acid were destroyed by gun-
                                                       fire in a remote area of the island.
                                                         The original plan called for all the waste to go to a single disposal
                                                       site. The shipping containers full of drummed waste would be un-
                                                       loaded from the ship, placed on trucks and  taken directly to the
                                                       disposal site. However, the site proposed  by the USEPA advised
                                                       the agency they could no longer take RCRA wastes,  only PCBs.
                                                       As a result, a substantial amount of the waste had to be recontain-
                                                       erized. Ultimately RCRA wastes were sent to  four sites in addition
                                                       to the one which took the PCBs. The delays resulting from recon-
                                                       tainerizing exceeded 96 hours, therefore requiring the team to get
                                                       temporary TSD permits issued to the transporter.

                                                       FIELD OPERATIONS

                                                         It  is beyond the scope of  this  paper to describe  in detail the
                                                       actions taken at all of the sites. A listing of the sites, and  the types
                                                       of waste materials removed,  is found in table 4. With  few excep-
                                                       tions, the sites were either pesticide/chemical storage areas or trans-
                                                       former storage  yards. However, detailed descriptions of opera-
                                                       tions at two sites,  good examples of the degree of flexibility re-
                                                       quired in a project of this complexity, follow this section.
                                                         The team used the same basic approach to all of the transformer
                                                       storage sites:
                                                       •Clear vegetation and debris from around transformers
                                                       •Clear a staging area if necessary
                                                       •Stage and number all transformers
                                                       •Open and sample all transformers for analysis
                                                       •Pump transformers with PCB concentrations between 50-500 ppm
                                                       •Pump transformers with PCB concentrations greater than 500
                                                        ppm and fill with diesel oil
                                                       •Pump contaminated diesel oil and rinse
                                                       •Pump transformers with PCB concentrations less than 50 ppm
                                                       •Prepare PCB transformers for shipment by partially filling  with
                                                        vermiculite,  reseating  the lid, wiping  down the exterior  and
                                                        collecting information off the data plate
                                                       •Load containers

                                                         The team used either an electric or gasoline driven pump, rein-
                                                       forced garden hose and a stinger fabricated from PVC water pipe
                                                       to drain transformers.  Some of  the larger transformers were
                                                       drained by connecting the pump to the drain  valve on the bottom.
                                                       The transformers shipped back to the mainland ranged in size and
                                                       volume from 110 Ib, 10 gal capacity to 30,000 Ib 1500 gal capacity.
 432
CASE HISTORIES

-------
Site/Location
            Table 4
Pacific Islands Project: Site Inventory

            Disposition
Guam
 1 Univ Guam Pest, site
 2 Univ Guam Ag Exp Stn
 3 Guam Mem Hosp lab
 4 PUAG Acid Waste site
 5 Deptof AgPest. site
 6 Baza Gardens drum site
 7 GEDAsite
 8 DPHSS Pest, site
 9 DPHSS Lab
10 Perez Bros drum site
11 Harmon Plaza site
12 Connell Bros drum site
13 GPA/Naval Stn PCB storage
A Deptof Ed lab
B delaCruz drum site
Saipan
14 CNMI PCB site
15 CNMI TTPI warehouse site
Palau
16 DP W PCB site
Majuro
17 PCB site
18 DeptofAgsite
19 DP W PCB site
20 DDT site
24 DPW old plant site
C Hospital lab
Ebeye
21 DPW PCB site
Eniwetok
22 DPW PCB site
Jaluit
23 DPW PCB site
Ponape
25 Electric shop
26 Nett Rd site
27 Deptof Forestry site
28 Gov't warehouse
Kosrae
29 DPW PCB site
Truk
30 DPW PCB site
31 Military dump site
Yap
32 DPW PCB/pest. site
F Pesticide site
G USCG PCB storage
Ejit
D Med Clinic pest, site
Kwajalein
E Army PCB storage
                                 Completed 2/29
                                 Completed 2/29
                                  Completed 3/1
                                  Completed 3/7
                                  Completed 3/6
                      Assess 3/8: No hazmat found
                                  Completed 3/1
                                  Completed 3/2
                                  Completed 3/7
                                RP cleanup 3/13
                                  Completed 3/1
                                RP cleanup 2/29
                                 Completed 3/13
                    Completed 2/29; no funds req'd
              Completed 3/8; funding $20K approved

                                 Completed 4/29
                                 Completed 4/29

                                 Completed 4/20

                                 Completed 3/27
                                 Completed 3/26
                                 Completed 3/23
                                 Completed 3/23
                     Assess 3/27: No hazmat found
                    Completed 3/22; no funds req'd

                                  Completed 4/4

                   DOE cleanup: No response req'd

                    Material moved: No resp. req'd

                                 Completed 4/14
                                 Completed 4/14
                                 Completed 4/14
                                 Completed 4/14

                                 Completed 4/13

                                 Completed 3/31
                     Assess 3/31: No hazmat found

                                 Completed 4/11
             Completed 4/11; funding $13K approved
                    Completed 4/11; no funds req'd

             Completed 3/24; funding $10K approved

                            Tech assist to RP only
 28 CERCLA funded cleanups (25 from orig. project + 3 new funded sites B, D, F)
 3 RP cleanups (Sites 10, 12, E)
 3 USEPA cleanup, no funds required (Sites A, C, O)
 5 5 Sites where no action required (Sites 6, 22, 23, 24, 31)
39 Total sites
  AH of the pumping operations involving PCB or PCB-contam-
inated oil or diesel oil were conducted by staff in Level C protec-
tive gear.
  The chemical  storage/pesticide storage sites were cleaned up
using one basic procedure. The work was conducted by a crew of
three or four persons in either Level C or B protection. In all cases,
the materials were inside laboratory storerooms or other confined
spaces too small for a crew of three or four to work effectively.
  At each site, the team set up a bench in front of the storage area.
Drums which would receive the waste were lined up on one side of
this  bench. One crew member removed waste materials from the
storage area  and brought them out to the bench.  A second crew
member recorded the chemical name or, if a pesticide, the active
ingredients of every container and verified the hazard class. The
remaining crew  members placed  the  materials  in a drum labeled
for the appropriate hazard class. Once all the materials were re-
moved from  the storage area, the interior spaces were decontam-
inated with chlorine bleach or TSP.
  One must  record the chemical  name or pesticide active ingred-
ients when making up a "lab  pack," because hazardous waste dis-
posal sites may not accept "lab packs" without it.

PUBLIC UTILITY AGENCY OF GUAM
ACID WASTE SITE
  The  Public Utility  Agency of  Guam (PUAG) acid waste site
located in Dededo, contained an estimated 200 15 gal polyethylene
lined drums  of  hydrofluorosilicic acid.  The drums were highly
deteriorated.  They all had cracked liners, and many had spilled all
or part of their contents onto the surrounding soil. Additional
waste materials  on this  site  included approximately 1000 Ib of
sodium silicon fluoride and quantities of assorted shelf chemicals.
  The  original cleanup plan  for this site called for the on-site
neutralization of the acid using locally available coral sand. This
plan was based on  the assumption that,  since the containers had
been exposed to the rainy tropical environment for several years,
the contents would have been diluted  considerably. This plan was
abandoned when field testing indicated that the acid had  a pH of
near 0 and a concentration of 4 to 6 n. The neutralization of such a
large quantity of strong acid on-site could cause many problems.
  The revised cleanup plan required the pumping of all acid into
polyethylene  drums which had to be obtained locally. The empty
containers were neutralized with a small quantity of coral sand and
disposed of at a local land fill dump. The contaminated  soil was
covered with clean coral sand. The other materials on-site  were
segregated into compatibility groups and packaged  in 55 gal open
head drums.  Decontamination water  and water used  for cooling
personnel on  this site was treated by filtration through a coral sand
berm.

EJIT ISLAND PESTICIDE STORAGE SITE
  During  the early stages of cleanup  actions on Majuro, the
USEPA was  informed of the  existence of a  pesticide  storage site
located on a neighboring island within the atoll. Ejit Island, located
a short boat ride to the northeast of Majuro, presently serves as a
relocation home for many of the former residents of Bikini Island.
The  island has no roads, no  electricity and no water or sewage
systems. Upon arrival on the  island, the OSC determined that the
pesticide storage room occupied the middle room of a small build-
ing; one end was a one-room schoolhouse, and the other end was a
part  time medical clinic. The   20 ft x 20 ft pesticide storage room
was  grossly contaminated by hundreds of bags of pesticides in a
highly deteriorated state.
  The cleanup of this site presented an unusual logistical  problem
since the island is not serviced by a commercial shipping line. In
order to ship  empty drums, safety equipment and other supplies to
Ejit, it was necessary to charter the services of a 16 ft aluminum
boat. Three round trips were required to transport the supplies and
the three man cleanup crew.
  The cleanup operation at the site required an entry team using
Level C protection and entailed the removal and containerization
of all pesticides followed by a chlorine bleach scrubbing of all in-
terior surfaces of the storeroom. Upon completion  of the removal
action, the team found that the filled drums of wastes were too
heavy to be  transported in the small boat.  Arrangements  were
made with the Marshallese Government which was able to provide
an old military landing craft  (LCU) to transport waste from Ejit
to the Port of Majuro.
                                                                                                   CASE HISTORIES
                                                                                                             433

-------
CONCLUSIONS

  With  completion of  the Pacific  Islands Project,  Region  9
USEPA successfully removed all of the hazardous waste it orig-
inally set out to remove, as well as additional waste brought to its
attention during the project.  The project was  also successful in
meeting time and budgetary goals. There are many factors which
made this complex project a success. The most important are:
•The investment of time into conducting a  thorough preliminary
 assessment, particularly where there was a lack of local capability
 to recognize,  assess and solve the problems associated with haz-
 ardous wastes.
                                                       •The extensive planning and front-end logistics effort which in-
                                                        sured that the supplies and equipment were delivered  on time, in
                                                        good condition and at the proper location. The prompt removal
                                                        and safe transportation of the wastes to a disposal site was also
                                                        assured.
                                                       •The use of multi-disciplinary teams set up to be self-sufficient.
                                                        They were given the flexibility and authority to deal  with unex-
                                                        pected situations which arose while in the field.
                                                       •The establishment and maintenance of good liaison with the local
                                                        governments involved. The valuable cooperation and  support re-
                                                        ceived might not have been provided without the established good
                                                        working relationship.
434
CASE HISTORIES

-------
RfcMUDlAL INVESTIGATION AND  FEASIBILITY  STUDY FOR
            THE  POLLUTION ABATEMENT  SERVICES SITE
                                 (OSWEGO,  NEW  YORK)

                                        DANIEL W. ROTHMAN
                                        JOHN C. GORTON, JR.
                                           URS Company, Inc.
                                            Buffalo, New York
                                          JAMES A. SANFORD
                      New York State Department of Environmental Conservation
                                            Albany, New York
INTRODUCTION

  The 15.6-acre Pollution Abatement Services, Inc. (PAS) site is
located near the eastern limit of the City of Oswego, New York, ap-
proximately 2000 ft south of Lake Ontario (Fig. 1). It lies within a
light commercial zone and is bounded on its south by East Seneca
Street and on its  remaining three sides  by  a designated wetland
formed along the stream channels of White and Wine Creeks. Ad-
jacent properties include a residence on the north, union hall on the
east, solid waste  transfer station (formerly the Oswego County
Landfill) on the south and radio station  on the west.
  From  1970  through  1977,  PAS  was  operated  as a  high-
temperature liquid chemical waste incinerator facility. Throughout
its active life, the facility experienced continuous operating pro-
blems, numerous air and water quality violations and mounting
public opposition. During this time, a large number of drums con-
taining various chemical wastes were collected and stored on-site,
as were liquid chemical wastes in several on-site lagoons. In 1977,
PAS was abandoned. It was subsequently listed among the top 10
priority sites on USEPA's initial National Priorities List.
  During the several years immediately following its abandonment,
a number of emergency remedial actions were taken at the site to
limit access,  contain hazardous materials and remove some of the
chemical waste products. During the summer of 1982, a major sur-
ficial  cleanup  of the  site was  undertaken, which included the
demolition and disposal of on-site facilities and the removal of ap-
proximately 8,000 drums and 80,000 gal  of liquid chemical wastes.
Immediately following this surficial cleanup, the present study of
PAS was commenced.
  The primary purpose of this study was to conduct a complete and
thorough site investigation in order to identify continuing sources
of contamination (especially subsurface sources); to define the ex-
isting geological,  hydrological and hydrogeological conditions at
the site; to identify the vertical and horizontal extent of contami-
nant migration; to assess the present  and potential impact of the
site upon the environment and human health; and to identify and
evaluate feasible remedial alternatives.
                         OSWEGO COUNTY LANDFILL
                       Figure 1
                     Location Map
 REMEDIAL INVESTIGATION
 Study Methods

  The following study methods were utilized to develop a site
 characterization upon which the  subsequent development  and
 evaluation of remedial alternatives were based:
 •Review and evaluation of existing data and information concern-
 ing the site at the outset of the study
•Development of Comprehensive Health/Safety and Quality As-
 surance/Quality Control Plans prior to the initiation of on-site
 activities
•Development of expanded site  base  map with  updated top-
 ography
•Performance of detailed geophysical studies of the site and sur-
 rounding area  using the  techniques  of  terrain conductivity,
 electrical resistivity and seismic refraction
                                                                                     CASE HISTORIES      435

-------
•Collection and analysis of surface  water and stream sediment
 samples from stations along White and Wine Creeks
•Installation of backhoe test pits and trenches, soil borings and
 groundwater monitoring wells on-site and in the surrounding area;
 performance of geophysical well logging and in situ permeability
 testing during the installation of the borings;  performance of
 slug tests in the monitoring wells after they had been developed
 and allowed to recover
•Performance of soil  screening tests  for total  volatile  organic
 compounds (VOC) and polychlorinated  biphenyls (PCBs)  using
 a mobile laboratory  on-site during  the installation of test pits,
 trenches and borings
•Collection of soil samples  from the test pits, trenches,  borings
 and monitoring wells for detailed laboratory analyses
•Collection of groundwater samples from each of the monitoring
 wells for detailed laboratory analyses
•Performance of follow-up detailed electrical resistivity survey in
 an area of the site which, on the basis of field  investigation re-
 sults, exhibited anomalous behavior
•Performance of preliminary aquatic and biological survey up-
 stream and downstream from the PAS site along White and  Wine
 Creeks

Site  Characterization

  The data obtained during the field investigation were collectively
analyzed to form an overall understanding of the PAS site and its
impact upon the surrounding area. This site characterization, as it
relates to potential remedial actions, is summarized in the following
paragraphs.
  The most important environmental feature of the PAS site is the
wetland  which borders the (approximately 7-acre) former active
area of the site along  White and Wine Creeks and extends north-
ward from the site along Wine Creek to Lake Ontario. A variety of
fauna  make use of the wetland area, including a wide variety of
avian species. Also, a  number of fish species appear to migrate up
White and Wine Creeks as far as the PAS site during the spring
spawning season. Almost all of the surface water flow from PAS
drains toward White Creek  on  the north and east. The drainage
area of this stream measures approximately 1340 acres upgradient
from PAS and increases by approximately 2% as it passes  through
the site.
  The  site  is  located  in a fairly  complex  geological setting
characterized by glacial deposition and reworking. Six stratigraphic
units were encountered  during the subsurface investigation, in-
cluding a surficial layer of fill material deposited on the site prior to
the construction of PAS. The two most important stratigraphic
features of the site are: (1) a dense, continuous lodgement till layer
which underlies the more permeable surficial soils; and (2) a depres-
sion in  the central area of the site which contains stratified
sediments to greater depths than found elsewhere on-site.  The site
is underlain by sandstone bedrock with low intrinsic permeability at
depths which range from approximately 30 to 50 ft.
  Separate groundwater  flow systems were observed  in the over-
burden soils, down to lodgement till and in the underlying bedrock.
The upper, unconfmed flow system has a water table configuration
which closely reflects surface topography. The water table gradient
in this system ranges from around 0.02 to 0.13 and slopes generally
northward toward Lake Ontario. The bedrock piezometric surface
is lower than the overburden water table and  also flows northward
with a gradient of approximately 0.004 to 0.007. Consideration of
the water table configuration and stratigraphy of the site  leads to
the conclusion that White and Wine  Creeks are effluent in nature
and  that these streams act as hydraulic  barriers which  intercept
groundwater flow through the surficial soils at PAS. This, in turn,
leads to  the secondary (and  most important) conclusion that sur-
face water, particularly White Creek,  is the most likely pathway for
contaminant  migration off-site. This conclusion is supported by
surface geophysical studies,  which detected no groundwater con-
taminant plume passing under White or Wine Creek from the PAS
site and by groundwater analytical data, which indicated that con-
                                                       taminants found in groundwater underlying the site are generally
                                                       absent on the opposite side of White Creek.
                                                          Surface water analytical data indicate high levels of contamina-
                                                       tion in the two on-site drainage ditches at PAS and detectable but
                                                       show relatively low  and non-persistent effects  of PAS  upon
                                                       downgradient water quality. However, the actual impact may have
                                                       been obscured to some degree by the short, winter sampling period
                                                       and by the  filtering effect of the marsh adjacent to the site. A
                                                       subsequent spring aquatic survey of White  and Wine Creeks in-
                                                       dicated that they are severely stressed at and  downstream from the
                                                       PAS site.
                                                          The soils and groundwater  underlying  PAS  are highly con-
                                                       taminated by a wide variety  of priority pollutants. Although these
                                                       subsurface contaminants are distributed  nonuniformly across the
                                                       site, a number of the most highly contaminated samples are located
                                                       near former chemical waste processing and  storage areas  such as
                                                       the lagoons, drum piles and buried subsurface tanks. Furthermore,
                                                       soil  screening  results clearly indicate that soil contamination
                                                       generally decreases with depth,  with most of the observed con-
                                                       tamination occurring within  the upper 10 ft of the soil. PCB-1248
                                                       was detected at relatively high concentrations in the area of the site
                                                       to the north and west of  the former drum loading platform (Fig. 2).
                                                       This fact, plus the visual observations  of a  black, oily substance
                                                       leaking from the dumping bin behind the former loading platform
                                                       (Fig. 3),  breaking out at the toe of slope downgradient from the
                                                       platform (Fig. 4) and breaking out on the ground surface above
                                                       subsurface Tank No. 9 (Fig. 5), indicate that there is an active
                                                       source (or sources) of PCB-contaminated liquid  leaking into the
                                                       subsurface in this area of the site.
                                                          The primary potential impact of the PAS site appears to be upon
                                                       surface-water-related activities at and downgradient from the site.
                                                       Groundwater resources north (downgradient) of PAS are little used
                                                       at present, with public  water available throughout the area, and
                                                       have little chance of being further developed  since much of th area
                                                                                                               ,\
                                                                                  Figure 2
                                                                  Existing Subsurface Tank and Drum Locations
 436
CASE HISTORIES

-------
                                                                                            Figure 4
                                                                        Contaminant Breakout Below Former Loading Platform
                          Figure 3
          Dumping Bin Behind Former Loading Platform
                           Figure 5
        Contaminant Breakout Above Subsurface Tank No. 9
north of PAS is undevelopable. Surface water, on the other hand,
is used not only by fauna inhabiting and migrating through the
wetland area, but also by fishermen on a seasonal basis.
  The off-site property  most directly affected by conditions or
future activities at PAS is the adjacent radio station; it receives con-
taminated surface runoff from the site, a portion of which will have
to be included in any regrading and capping of PAS. Consideration
was also given to the potential impact which the adjacent Oswego
County Landfill (OCL) might have upon the PAS site. Although
the OCL is upgradient from, and in direct hydraulic contact with
PAS,  the quality of surface water and groundwater at the two sites
are distinctly different,  with  no  evidence  of hazardous  con-
taminants originating from the OCL.

EVALUATION OF REMEDIAL ACTIVITIES
Methodology

  Initially, individual remedial measures considered practical at the
site were identified and  expanded specifically for application at
PAS.  Next,  these  individual  measures  were incorporated, in
various combinations, into six alternative remedial plans. Each of
the plans was rated using a weighted matrix system which "scored"
the plans on the basis of four categories: level of cleanup, reliabil-
ity, longevity and operation/maintenance requirements. The cost
of each  alternative  was  computed as  a basis for  comparison,
although  the final selection of a remedial alternative was based
primarily upon its overall effectiveness.
Identification of Individual Measures
  Initially, the "universe" of potential remedial measures at hazar-
dous waste sites was narrowed for specific application to PAS on
the basis of technological, practical and preliminary cost considera-
tions.  Conceptual designs and  preliminary cost estimate were
developed for each of the  individual measures which survived this
screening process. A conceptual design and cost estimate, respec-
tively, for one of the measures which is considered to be feasible at
PAS—a shallow leachate collection system—are shown in Figure 6
and Table 1. In all, the following measures were considered:
•Limited Excavation and Removal—Although  wholesale removal
 of all contaminated soil from the site was not considered to be
 practical, the excavation  and removal of the  remaining  subsur-
 face drums and storage tanks (Fig. 2) was evaluated.
•Grading and Capping—The poor surface drainage and permeable
 nature of the soil cover at PAS  suggest the benefits of regrading
 and capping the site. The following capping/cover  systems were
 initially considered:  layered soil  covers,  bentonite  admixtures
 and membrane  liners. At this site, combined effectiveness and
 cost considerations led to the selection of a 50-mil high density
 polyethylene (HDPE) membrane cap.
•Stream Diversion—The existing topography and drainage char-
 acteristics at PAS would permit the rerouting of White Creek,
 which presently flows adjacent to and collects leachate from the
 most contaminated areas  of the  site. The option of diverting this
 stream to Wine Creek via a 60-in. diameter conduit, approxi-
 mately 1,000 ft in length, was considered.
                                                                                                  CASE HISTORIES
                                                          437

-------
'Slurry  Trenching—The construction of a slurry wall was evalu-
 ated from two aspects—as a perimeter wall for containment of
 contaminated groundwater and  as an upgradient groundwater
 diversion. The perimeter containment wall, extending over ap-
 proximately 0.5 mile to an average depth of 14 ft, was considered
 to be most practical and was carried  forward  in the evaluation
 process.
•Leachate Collection—Collection  of shallow leachate, via a series
 of collection  drains discharging  to a  duplex pump station and
 thence to an existing on-site storage  tank, was evaluated (Fig.
 6 and Table 1).
•Groundwater Recovery—An existing stratigraphic depression,
 containing highly  contaminated  sediments,  is  located near the
 center  of the site. Groundwater recovery from this depression
 was evaluated, again using the existing on-site tank for temporary
 storage prior to leachate disposal.
•Leachate Disposal—Two methods of leachate disposal were con-
 sidered—on-site treatment and off-site removal for treatment by
 a private facility. The anticipated quantity of leachate to be dis-
 posed  of  led to the selection of on-site treatment.  Based upon
 analytical results obtained during the study, but pending a bench-
 scale treatability study, a treatment  process consisting  of flow
 equalization, precipitation/flocculation/sedimentation, activated
 carbon adsorption and neutralization was preliminarily selected.
•Miscellaneous Items—In  addition to the previous list, several
 miscellaneous, non-construction  measures were evaluated. These
 include a detailed environmental assessment of the stream/wet-
 land system downgradient from  PAS and various regulatory ac-
 tions. Since contamination from  the site is continuing, a progres-
 sive remedial program is warranted, with the initial step being to
 control the source of contamination through implementation of
 (some  combination of) the remedial measures  listed above. The
 purpose of an environmental assessment would be to determine
 whether additional off-site corrective  measures will be required
 at  a later date. The regulatory actions evaluated  include: re-
 quiring disconnection of the few remaining drinking wells down-
 gradient   from  PAS and connection  to  the  already-available
 public water system; institution  of a temporary fishing  ban for
 the stream downgradient from the site; and placement of a tem-
 porary moratorium on new development in a small area immedi-
 ately adjacent to the Smith's  Beach marsh (Fig.  1). The latter
 two regulatory actions would be temporary, pending the outcome
 of the environmenal assessment. Furthermore, disconnection of
 the potable water wells is  considered to be conservative and pre-
 cautionary since there is no evidence at present to indicate con-
 tamination of these downgradient wells.

                           Table 1
              Cod Estimate for Leachate Collection
                                                                            PROPOSED LEACHATE
                                                                               PUMP STATION
DESCRIPTION
                                                COST
CapfUI Cort*
Leachate Drain (430 ft. installed, Including perforated PVC,
stone filter and filter fabric wrapping)
Leachate Collector! (850 ft)
Pump Station (installed, including itainleu iteel duplex unit,
72- in. fiberglass basin, controls and electrical service)
Forcemain (420 ft., installed, 2.5 in PVQ
Subtotal/Estimated Construction Costs
Engineering/Legal/Admlniitrative (15ft)
Contingency (20ft)
Total Estimated Capital Cost
Operation and Maintenance Costs (Annual)
Labor and Materials
Electricity
TOTAL ANNUAL O&M COSTS
Present Worth of O&M Costs (5 yrs., 7 5/8ft)
TOTAL PRESENT WORTH
$ 3.100
S 9,300
38,800
1,400
$52,600
7,900
10,500
17 1.000
300
50
$ 350
1,400
$72,400
                                                                                  Figure 6
                                                                      Conceptual Leachate Collection Plan

                                                       Development and Evaluation of
                                                       Remedial Alternatives

                                                         Of the individual remedial measures discussed in the previous
                                                       section, some are independent of each other, some are complimen-
                                                       tary and others are mutually exclusive. These functional relation-
                                                       ships were used to incorporate the individual measures, in various
                                                       combinations, into six alternative remedial plans. The plans are
                                                       identified in Table 2.
                                                         Following plan development, a system was developed and util-
                                                       ized to evaluate the relative effectiveness of the plans. This evalua-
                                                       tion procedure consists of a  weighted matrix with each alternative
                                                       rated individually in the following four categories:
                                                       •Level of Cleanup
                                                       •Reliability
                                                       •Longevity
                                                       •Operation and Maintenance Requirements
                                                         The categories are scored independently on a "1 to 5" scale, with
                                                       "1" representing the lowest  (or worst) score and "5" representing
                                                       the highest (or best). The independence of the scoring system is im-
                                                       portant. For example, an alternative which initially provides a high
                                                       level  of cleanup but  has  a  limited  or uncertain life expectancy
                                                       would score  high in  the  former category but  low in  terms of
                                                       "longevity." The  weights of the various categories are also as-
                                                       signed  independently  and  partially reflect  the attitudes  and
                                                       preferences which had been expressed by local officials, regulatory
                                                       agencies and the general public.
                                                         The scoring system developed and used in this study is only semi-
                                                       quantitative.  Unlike costs, the incremental benefits of alternative
                                                       plans cannot be quantified on an absolute basis. Rather, the scor-
                                                       ing system provides a relative basis for comparison. Although costs
                                                       have been calculated for each of the alternative plans, primary em-
                                                       phasis in the evaluation procedure has  been placed  upon their
                                                       relative effectiveness.  The scoring of the  six alternative remedial
                                                       plans evaluated at PAS is given in Table 3.
                                                       Recommendations

                                                         Alternative No.  1 is the most effective and the most costly of the
                                                       six  plans evaluated  (Table 2).  However,  the other  remedial
                                                       measures (e.g., excavation and removal of the entire contaminated
                                                       surficial soil layer) are more  expensive than Alternative No. 1, but
438
CASE HISTORIES

-------
Alternative
                 Table 2
Alternative Remedial Plans for the PAS Site

        Components
                                                                                                    Table 3
                                                                                    Evaluation of Alternative Remedial Plans
Alternative No. 1        Limited Excavation and Removal
                     Grading and Capping
                     Perimeter Slurry Wall to Lodgement Till
                     Leachate Collection
                     Groundwater Recovery
                     On-site Treatment
                     Miscellaneous Items

Alternative No. 2        Limited Excavation and Removal
                     Grading and Capping
                     Stream Diversion
                     Leachate Collection
                     Groundwater Recovery
                     On-site Treatment
                     Miscellaneous Items

Alternative No. 3        Limited Excavation and Removal
                     Grading and Capping
                     Stream Diversion
                     Leachate Collection
                     On-site Treatment
                     Miscellaneous Items

Alternative No. 4        Limited Excavation and Removal
                     Grading and Capping
                     Stream Diversion
                     Miscellaneous Items

Alternative No. 5        Limited Excavation and Removal
                     Grading and Capping
                     Perimeter Slurry Wall to Lodgement Till
                     Miscellaneous Items

Alternative No. 6        Limited Excavation and Removal
                     Grading and Capping
                     Miscellaneous Items
 were eliminated during the preliminary screening process for one or
 more of the reasons mentioned in the evaluation section. Alter-
 native No . 1 alone provides a containment slurry wall and ground-
 water recovery system, thereby addressing  the possibility of  a
 groundwater  contaminant   migration  pathway   from   the
 stratigraphic depression in the central area of the site. Although the
 remedial investigation did  not indicate  the presence of such  a
 pathway, its existence could not be ruled out. On this basis, and
                                                                                                Alternative Nos.

                                                                                              234
                                                            Level of Cleanup
                                                            (Weight = 3)
                                                             Category score
                                                             Weighted score
                                                            Reliability (Weight = 1)
                                                             Category score
                                                             Weighted score
                                                            Longevity (Weight = 1)
                                                             Category score
                                                             Weighted score
                                                            O&M Requirements
                                                            (Weight = 1)
                                                             Category score
                                                             Weighted score

                                                            Total Weighted Score

                                                            Total Estimated Cost
                                                             ($ X  103)
 5
15
 2
 2

24
 5
15

 3
 3

 3
 3
 2
 2

23
 4
12
 4
 4

23
 5
 5

20
 5
 5

21
 5
 5

15
                                                                                     2,281.9 2,166.9 2,050.7 1,442.0  1,557.0  1,219.5
                                                          pursuant to conversations with the New York State Department of
                                                          Environmental Conservation and the USEPA, Alternative No. 1
                                                          was recommended. In summary, this recommended alternative in-
                                                          cludes the  following  measures: limited excavation and removal of
                                                          subsurface tanks and drums; construction of a perimeter slurry wall
                                                          around the  site; construction  of a  shallow leachate  collection
                                                          system; installation of a groundwater recovery system; construction
                                                          of an on-site treatment facility; grading and capping of the site; and
                                                          performance of an off-site study and institution of various regu-
                                                          latory actions.
                                                            The recommended remedial plan for PAS will cost an estimated
                                                          $2.28 million (total  including  additional  study,  design and  con-
                                                          struction) and require approximately 68 weeks to fully implement.
                                                          Permits  will be  required for  wetland disturbance and discharge
                                                          from the on-site treatment plant. A long-term baseline and post-
                                                          closure monitoring program has been recommended to accompany
                                                          the plan.
                                                                                                         CASE HISTORIES
                                                                                                                                    439

-------
         CASE STUDIES  INVOLVING THE  TREATMENT  OF
     HAZARDOUS  SUBSTANCES  UNDER  THE SUPERFUND
                           REMEDIAL ACTION PROGRAM

                                          WILLIAM  M.  KASCHAK
                                          JAMES J. SPATARELLA
                                   U.S. Environmental Protection Agency
                                       Hazardous Site Control Division
                                               Washington,  D.C.
INTRODUCTION
  The USEPA began the Superfund cleanup program after passage
of CERCLA in 1980. The Superfund program operates under the
guidelines of the  National Contingency Plan (NCP), which was
published on July  16, 1982. This plan was expanded by the USEPA
to provide new Federal authority to respond to the  problems at
abandoned or uncontrolled hazardous waste disposal sites. The
NCP  outlines three categories of response actions: immediate
removals, planned removals and remedial response actions.
  Remedial response actions are intended to achieve solutions con-
sistent with permanent remedy at uncontrolled hazardous waste
disposal sites. As such, more time and effort are required to deter-
mine  the "appropriate extent of remedy—the  least  expensive
remedy that is technologically feasible and reliable, effectively
reduces the danger and adequately protects public health, welfare,
and the environment". The NCP identifies three types of remedial
response actions which are based  upon the complexity, immediacy
and extent of the hazards:  (1) initial remedial measures, (2) source-
control and (3) off-site remedial actions.
  Initial remedial  measures are appropriate when the actions to be
taken  are limited  in nature and require a minimum of planning.
The source-control and off-site remedial actions are more complex
and require more extensive engineering evaluations to select the
most cost-effective solutions.
  An initial remedial measure is being implemented at the Bridge-
port Rental and Oil Services (BROS) site in New Jersey involving
the  treatment and disposal of  the aqueous phase of  an  11.8-acre
lagoon. Source-control remedial action is being implemented at the
Sylvester site in New Hampshire in two phases. The first phase in-
volved the installation  of a slurry wall and cap. The second phase
will include the extraction and treatment of highly contaminated
groundwater within the containment system. The  planning ac-
tivities and engineering studies leading up to the selection of the
treatment systems at these two sites are discussed below.

BRIDGEPORT RENTAL AND OIL SERVICES

  The 26 acre BROS site is a former oil processing and reclamation
facility located in Bridgeport,  New  Jersey. The  predominant
feature on the site is an unlined 11.8-acre lagoon averaging 12 to 18
feet in depth with  the greatest depth at 60 ft. A thick layer of heavy
oils laced with construction debris, drums and other trash floats on
the  surface. There are  also several large, partially submerged tank
trucks in the lagoon. Some 80 storage tanks and vessels ranging in
capacity from a few thousand  gallons to greater than 300,000 gal
are also present. The majority of the tanks are either empty or con-
                                                   tain bottom sludges with only two of the larger tanks containing
                                                   substantial quantities of liquids.
                                                     There is an eight to ten acre area of land adjacent to the lagoon
                                                   with stressed vegetation. This damage occurred when the dike sur-
                                                   rounding  the lagoon was breached, spilling some of the lagoon's
                                                   contents. There are visual seeps around the lagoon and into the ad-
                                                   joining freshwater ponds.  Waterfowl are constant victims as they
                                                   attempt to land on the lagoon surface.
                                                     The bottom of the lagoon is unlined; however, a thick layer of
                                                   oily sludge on the bottom retards exfiltration. By June  1982, the
                                                   level of the lagoon had risen to within 6 in. of the top of the dike.
                                                   This situation required an immediate removal action. This action
                                                   included mobilization of the USEPA's transportable activated car-
                                                   bon unit. This mobile treatment system  was used to lower the
                                                   lagoon level by approximately 2  ft,  thus developing  adequate
                                                   freeboard. This action  involved  the  removal,  treatment  and
                                                   discharge  of approximately  5,000,000  gal of treated  water to
                                                   Timber Creek.
                                                     Camp,  Dresser and McKee, Inc. (COM) initiated a Remedial In-
                                                   vestigation' in the Fall of 1981 to determine the extent and severity
                                                   of contamination at the site. Concurrent  with the emergency ac-
                                                   tion, the scope of this effort was redirected to determine the most
                                                   cost-effective method to lower the lagoon level even further. Upon
                                                   completion of this effort, the USEPA and the State of New Jersey
                                                   entered into a Superfund State Contract on Oct. 29, 1982, to design
                                                   and implement the initial remedial measures at the site. The objec-
                                                   tive of the initial remedial measure was to reduce the liquid level in
                                                   the lagoon to ensure that overflow of the liquid over  the dikes
                                                   would be delayed for a substantial period of time while alternatives
                                                   for the long term remedial action were being evaluated.
                                                     To  determine the  most  cost-effective  approach to the initial
                                                   remedial measure, sampling was begun to characterize the contents
                                                   of the lagoon. Samples of the lagoon oil and  aqueous phase were
                                                   taken  at several locations and various depths and were analyzed at
                                                   one of the USEPA contractor laboratories. Some limited data were
                                                   available  from the testing during the operation of the  USEPA's
                                                   mobile treatment system. These chemical data were used as a basis
                                                   for designing initial treatability studies  of the  liquid wastes.
                                                     The general characteristics of the liquid  wastes were: TOC,
                                                   180-220 mg/1; COD,  720 mg/1; five-day BOD, 90 mg/1; TSS, 690
                                                   mg/1;  VSS, 300 mg/1; and oil and grease,  80 mg/1. In general, the
                                                   waste  is composed of approximately one-third volatile organics,
                                                   one-third  large molecular weight oily-type materials and the re-
                                                   maining one-third  is generally uncharacterized and consists of a
                                                   variety of organic compounds. The organic species found in signifi-
                                                   cant quantities were benzene, trans - 1-2, dichloroethene, methy-
440
CASE HISTORIES

-------
                                                                       I  I POLYMER
                                                               I—^>-U
                                                H2SO.
                  D-C-!
                  DE-EMULSIFIER
                                                                                                       TOORANULAR
                                                                                                       ACTIVATED
                                                                                                       CARBON SYSTEM
                                                                                   •o  FLOW HATE

                                                                                      TOTAL ORQANIC CARBON

                                                                                      SAMPLING POINT
                                                           Figure 1
                                              Treatment Process Schematic Bridgeport
 lene chloride and toluene, all having concentrations around or
 above 1 mg/1. The inorganic substances of concern are lead and
 zinc,  while most of the remaining metals have relatively low con-
 centrations.
 Oil Layer

  The oily layer of the lagoon had a very high viscosity. The oily
 layer  actually moves about the  surface of the lagoon depending
 upon the wind direction. It contained levels of PCBs close to 450
 mg/1. A wide variation of metal species were present in the oil with
 a significant difference in concentrations between the organic liq-
 uids and aqueous samples.  Several metals concentrations  were,
 however, found in the 1-10 mg/1 range.  A major element of any
 remedial actions taken at  the site will involve  dealing with the
 physical and chemical characteristics of the oily layer.
  The level of the  lagoon  could  have  been lowered either by
 treating the underlying aqueous phase or by treating a combination
 of the oily surface layer and the  aqueous phase. Although the oily
 surface layer represented a  significant threat to the environment, it
 was determined that it could be more appropriately addressed dur-
 ing the long term remedial  action rather  than  under the initial
 remedial measure because of difficulties of handling and disposal.
 Moreover,total  removal of the oily layer would only reduce the
 level of the lagoon by 1 to 2 ft.
  A work group consisting of personnel  from the USEPA, CDM
 and the New Jersey Department of Environmental Protection was
 formed to make all decisions on the technical aspects of the project.
 The group  decided that the initial remedial measure would be to
 treat the aqueous phase only. The work group also decided that the
 level of the lagoon should only be lowered  to the level of the sur-
 rounding groundwater to maintain  hydraulic equilibrium between
 liquid levels and lessen the  possibility of  breaching the  bottom
 "seal," which was a result of the bottom sludges. This would re-
 quire lowering the liquid level by approximately 12 ft, or treating
 approximately 35,000,000 gal of the aqueous  phase of the lagoon.
  The work groups established discharge limits for Total Organic
Carbon (TOC) for the interim treatment system. TOC was used
because of the multitude of organic compounds present at the site.
Since most  of these compounds  were present at concentrations in
the  50-100  /tg/1  range,  monitoring for TOC would markedly
simplify the analytical process. The discharge limits agreed upon
for TOC were 50 mg/1 daily average over a 30-day period with a
daily maximum of 100 mg/1.
  The recommended wastewater treatment process was based upon
the raw water quality data, effluent discharge limits and several
bench scale and pilot tests.  Final design criteria were established
during the design project. A process diagram is provided in Figure 1
with the unit processes described in the next section.

Wastewater System Design

  Oil Separation. The removal of oil at the oeginning of the treat-
ment process would protect  subsequent processes from any heavy
concentrations of emulsified, non-aqueous organics. The oil treat-
ment process consists of addition of a demulsifier followed by a set-
tling to allow the oil  to float  to the surface for  collection and
removal.
  Air-Stripping. Air-Stripping was recommended for  the removal
of most of the volatile organics to reduce organic vapor levels near
open process tanks and to maximize the life of the carbon adsorp-
tion units.  Laboratory studies indicated that air-stripping would
reduce the  TOC by 50-80 mg/1. A countercurrent packed tower
with an air-to-water flow ratio of 50:1 was recommended.
  Flocculation/Sedimentation.  Flocculation  and  sedimentation
were recommended prior  to carbon adsorption. The process in-
volves coagulant addition and pH adjustment to remove suspended
solids, oil and inorganic compounds. A flash mixer,  flocculation
basin and   sedimentation  basin  are  used.  Suspended solids
discharges for this unit process are 30 mg/1. Bench scale studies
were performed to  select the coagulants and dosages for proper
operation.
  Granular Activated  Carbon (GAC).  A GAC unit was  recom-
mended as  the most effective method of removing the remaining
organic  compounds. The recommended process includes two dual-
column  adsorption modules operating  in parallel. Each of the
parallel  systems consists of two carbon beds operated in series; each
column  has an empty bed contact time of  30 min.  Information
from the operation of the USEPA's mobile carbon treatment
                                                                                                  CASE HISTORIES
                                                          441

-------
                           Table 1
        Concentration of Organic Contaminants Found In the
                Groundwater at the Sylvester Site


POLLITTANT
Vinyl Chloride
Benzene
Chloroform
1,1, 2-Tr icholoroe thane
Ethylene Chloride
Tetrachloroethylene
Trichloroethylene
Xylenes
Methyl isobutyl Ketone
Methyl Ethyl Ketone
Chlorobenzene
Methylene Chloride
Toluene
Ethyl Benzene
1 , 1-Dichloroethane
t-1 , 1-Dichloroe thane
1,1, 1-Trichloroethane
Methyl Methacrylate
Ethyl Chloride
Te t rahyd ro t ur an
2-Butanol
Dimethyl Sulfide
Diethyl Ether
Methyl Acetate
Isopropyl Alchol
Acetone
HIGWST CQNC.
FOUND IN
GROUND WATER
(PPB)
950
3,400
31,000
17
73,000
570
15,000
10,000
21,000
80,000
1,100
122,500
29,000
1,200
15
18,000
2,000
3,500
320
1,500,000
3,560
3,500
20,000
2,400
26,000
310,000
system and bench scale studies was used as a basis for the selection
of GAC.
  Equalization Basin. A holding tank was recommended at the end
of the treatment system to ensure the quality of the effluent before
its discharge to Timber Creek. In  the event that the quality of the
effluent did not meet the discharge limits, the effluent would be
returned to the lagoon.
                                                          Disposal of Process Sidestreams. It was recommended that the
                                                        side streams be returned to the lagoon. These recycle streams in-
                                                        clude collected oil, backwash from the carbon units, sludge from
                                                        the sedimentation basin and other miscellaneous streams such as
                                                        personnel  and equipment decontamination  water. The spent car-
                                                        bon could either be regenerated or disposed off-site.
                                                          The design of the interim treatment was completed by CDM. The
                                                        construction and operation is being accomplished by the U.S. Ar-
                                                        my Corps of Engineers. The treatment plant was fabricated and put
                                                        into operation in December  1983; however, the system was shut
                                                        down during the winter months. Operation of the  treatment plant
                                                        has resumed.
                                                        SYLVESTER HAZARDOUS WASTE SITE

                                                          The Sylvester site is a six-acre site in Nashua, New Hampshire,
                                                        which was originally used as a  sand borrow pit for a number of
                                                        years. During the late 1960s, the operator of the pit began an unap-
                                                        proved  and illegal waste disposal operation, apparently intending
                                                        to fill the  excavation. Household refuse,  demolition  materials,
                                                        chemical sludges and hazardous liquid chemicals were all dumped
                                                        at the site  at various times. The household refuse  and demolition
                                                        materials were usually buried, while the sludges and  hazardous li-
                                                        quids were either mixed with the trash or were  allowed to percolate
                                                        into the ground adjacent to the sand pit. Some hazardous liquids
                                                        were also stored in drums which were either buried or left on-site.
                                                        While it is impossible  to estimate the total  quantities of waste
                                                        materials discarded at the site, the USEPA has documented that
                                                        over 800,000 gal of hazardous waste were discarded there during a
                                                        ten month period in 1979.
                                                        Groundwater Contamination

                                                          A contaminated groundwater plume is  moving from the site
                                                        toward  Lyle Reed Brook. The uncontrolled  plume had the poten-
                                                        tial of contaminating all private drinking water wells between the
                                                        site and  the Nashua River as well  as becoming a major source of
                                                        stress on the Nashua River. In addition, Lyle Reed Brook would
                                                        not be able to support any aquatic life and would pose a direct
                                                        threat to human health at the adjacent trailer park  from volatiliza-
                                                        tion of the organic pollutants in the brook.
                                                          The USEPA used CERCLA emergency funds to offset the threat
                                                        by installing a groundwater interception and recirculation system at
                                                           Table 2
                                              Suitability of Treatment Processes at the
                                                  Sylvester Site, New Hampshire
UNIT
OPERATIONS
Air Stripping
Stream Stripping

Carton Adsorption

UV-Ozone


Biological

Vtet Oxidation

Reverse Oarosis
pH Adjustment
Precipitation
VOLATILE
ORGANICS
Inadequate ffenoval
Effective Concentrated
Technique
Inadequate hotn^jl

Nit suitable IXie to
High Concwntr.it ion

Effective l»rn ,v.,i
Technique
N3t Suitable a«- to ij~
Concentration
Inadequate Tr.'atnn_mt
Not Applicable
NOW- VOLATILE
CtCANICS
N-K Suitable
Not Suitable

Effective Ronoval
Te.-hnique
NX Suitable Due
to High Concentration

Effective ftenvjval
Technique
Wit Suitable Due to
Low Concentration
Nit Applicable
NH Applicable

INORGANICS
K>t Suitable
Not Suitable

Not Suitable

Hay oxidin and
precipitate sen*
metals
Nat suitable «etv. s
Ton ic
N3t Suitable

Difficult Cperation
Effective Ronoval
Technology
             Electrodialysis


             Ion Exchange
                             Not Applicable


                             N-.t Applicable
Not Applicable


NH Applicable
Inefficient Cperation/
Inadequate taioval

Inappropriate Technology-
Difficult Operation
442
CASE HISTORIES

-------
                                       COAGULANT AID
                                                                                         SULFURIC ACID
                                                                                             AIR •LOWERS
                                                           Figure 2
                                                Process Schematic, Sylvester Site
the site in November, 1981.  The purpose  of the system was to
retard further migration of the plume until a remedial action could
be implemented. The system operated until October, 1982, when it
was replaced by the first phase of remedial  action.
  The New Hampshire Water Supply and Pollution Control Com-
mission has been directing efforts to contain and clean up the site
using a two-phased approach for managing contamination. Their
approach was developed as a result of the Remedial Investigation
and the Feasibility Study2 completed by Roy F. Weston, Inc. in
January,  1982. The project design was reviewed,  approved  and
funded by the USEPA in  July, 1982,  as the most cost-effective
remedial  alternative  that  adequately  protected human health,
welfare and the environment.
  The first phase of the remedial action involved containment of
the contaminated plume. This step required the installation of a
bentonite slurry wall reaching depths of up to 90 ft in order to key it
into bedrock.  Then the entire 20-acre area was covered by an im-
permeable surface cap. The immediate  purpose of the cap was to
temporarily contain  the  contaminated  groundwater  while  the
groundwater treatment plant was designed  and  built. The second
purpose of the installing the slurry wall and cap was to exclude
clean water from entering the  contaminated site once groundwater
treatment  was initiated.  The implementation of this phase  was
completed in October, 1982.
  The second  phase of the remedial action was to design and build
a groundwater treatment system capable of reducing the ground-
water contamination to an acceptable level. The initial treatability
work, completed as part of the feasibility study, was comprised of
bench scale treatability studies using representative groundwater
samples.
  The Remedial Investigations  showed that there were high con-
centrations of heavy metals as well as volatile and extractable
organic concentrations in the groundwater under the site (Table 1).
No one unit  operation is  capable of  removing all of the con-
taminants. The treatability studies evaluated the potential of dif-
ferent unit operations to adequately remove particular groups of
contaminants. The development of an appropriate treatment train
capable of effectively removing all contaminant groups  is  il-
lustrated in Table 2.
  The  data clearly show that only two treatment trains will ade-
quately treat the contaminated groundwater to the required levels.
Both treatment  trains  required the initial  removal of inorganic
materials by  chemical  neutralization and precipitation methods.
This requirement is the result of the high concentrations of iron and
manganese (averaging 350 mg/1  and 80 mg/1 respectively) which
precipitate out of solution in any process that introduces air in to
the groundwater. The  introduction of air  into the groundwater
results  in the  plugging or fouling of the organic treatment system.
  The  first suitable treatment train uses steam stripping  to remove
volatile organics while the second train uses biological methods.
The feasibility study estimated the minimum treatment rate  for
both trains to be 35 gal/min, 24 hr/day continuous treatment, and
the optimum  treatment rate would be 100 gal/min. Upon the com-
pletion of a supplement to the feasibility study3 in July 1982,  the
optimum treatment rate was later revised to 300 gal/min con-
tinuously. The supplement provided additional information about
the costs associated with various groundwater treatment  rates. This
increase in the  rate of treatment reduced the expected treatment
time from 6.2 to 1.7 years. This period is based on the estimated ef-
fectiveness of a 90% reduction of all contaminants from two full
flushes of the contained volume.
  The  operation of a pilot plant was necessary prior to the design
of the  full-scale plant because of the wide variations in the concen-
tration of individual pollutants. The design for the pilot plant was
completed  in November 1982, the unit was built on the Sylvester
site and experiments  were begun  in February, 1983. The plant,
designed to  remove inorganics  and volatile organics from  the
groundwater  that was pumped from three on-site wells, is shown in
Figure 2.
  The  inorganic  chemical treatment process is designed for  the
removal of iron  and manganese. This process consists of chemical
precipitation  of heavy metals, pH adjustment of the wastewater
                                                                                                  CASE HISTORIES
                                                          443

-------
and sand filtration to remove the precipitated metals.  In the pilot
tests, iron removal was greater than 99% under all conditions, and
manganese removal was greater than 99.8% at pH's of 10 and 11,
with polymer doses between 0.5 and 1.0 mg/1.
  The  next process step is removal of the volatile organic com-
pounds using a High Temperature Air Stripper (HTAS). The con-
taminated groundwater (with the  metals removed) is preheated in
two heat exchangers, one an economizer and the second a trim heat
exchanger, before entering the column in which steam strips off (he
organics. Over the range of operating temperatures tested, all of
the priority pollutants and more  than 75%  of the alcohols  were
removed from the wastewater.
  In addition  to  the  above  pilot  tests,  bench and  pilot  scale
treatability studies were conducted on distillation, incineration and
biological treatment systems.  The distillation studies  were  con-
ducted by Artisan Industries, Inc.  The purpose of their studies was
to determine  the  feasibility  of concentrating the  organic  con-
taminants present in  the  HTAs condensate to 50-60% organics
while leaving the bottoms free of volatile organics. The data in-
dicate that it is technically feasible to make a reasonable separation
of the  volatile organics from the HTAS condensate by standard
distillation techniques. The limitations of this unit process are the
need for a feed liquid containing only very small amounts of non-
condensible gases (i.e., air) and the need for a  vapor phase ac-
tivated carbon solvent recovery system to remove the remaining sol-
vent vapors from the condensers. These requirements increased the
final cost of this process much above the incineration alternative.
  The incineration studies were conducted by Trane Thermal, Inc.
The purpose of these studies  was to determine the optimum in-
cineration design for the destruction of the aqueous waste  and to
determine the amount of fuel required per pound of  waste. The
aqueous wastes were generated in the pilot plant by passing the ef-
fluent air stream from the HTAS through a condenser. A destruc-
tion efficiency of tetrahydrofuran for all three test runs greater
than 99.98% was attainable. The quantity of tetrahydrofuran col-
lected was below detectable limits of 1.5 jig. The tests indicate that
                                                        a residence time of 1.0 sec at 1500 °F temperature will be sufficient
                                                        for achieving a 99.99% destruction efficiency of tetrahydrofuran,
                                                        the principal organic hazardous constituent.
                                                           Hydrocarbon emissions were also evaluated as a function of in-
                                                        cineration temperature for  optimum conditions. The studies in-
                                                        dicated that the presence of aqueous waste did not increase the
                                                        hydrocarbon  emissions.  Therefore,  almost  all the  unburned
                                                        hydrocarbons were the result of incomplete combustion of the fuel.
                                                        Since the heating value of the aqueous waste was determined to be
                                                        less than 0.5%, the amount of fuel required  per pound of waste
                                                        varied from 5114 BTU at 1450°F to 6910 BTU at 1600°F.
                                                           The biological treatment studies were conducted by Environmen-
                                                        tal Engineering Consultants, Inc.  The purpose of these tests was to
                                                        establish the  treatment  efficiency of activated sludge, aerated
                                                        lagoon and rotating biological contactor systems on the HTAS ef-
                                                        fluent. All three systems operated at high levels of efficiency in
                                                        removing the total  phenols and the extractable organics of concern.
                                                        The operational differences between the three systems were not
                                                        significant enough  to eliminate any system. Therefore the activated
                                                        sludge system was  chosen on the  basis of the economic analysis.
                                                           A Supplemental Record of Decision on the groundwater treat-
                                                        ment alternative selection was  approved in  September 1983. The
                                                        construction contract  was awarded in April 1984 and will be com-
                                                        pleted by the Fall of 1984.

                                                        REFERENCES
                                                        1.  Camp, Dresser & McKee, Inc., Initial Remedial Action at the Bridge-
                                                           port Rental A Oil Services Site,  \ J.. Feb.. 1983.
                                                        2.  R.F. Weston, Inc., Final Report-Sylvester Hazardous  Waste Dump
                                                           Site Containment  and Cleanup Assessment,  prepared  for the  New
                                                           Hampshire  Water Supply and  Pollution  Control Commission,  Jan.
                                                           1982.
                                                        3.  Roy F. Weston, Inc., Supplemental Study to Final Report on Sylvester
                                                           Hazardous  Waste Dump Site Containment and Cleanup Assessment,
                                                           prepared  for the  New Hampshire Water Supply and  Pollution Con-
                                                           trol Commission,  July 1982.
444
CASE HISTORIES

-------
           CLEANUP  OF  RADIUM PROCESSING RESIDUES
                          IN A HIGHLY URBANIZED AREA

                                                JOHN M. BRINK
                            U.S. Environmental Protection Agency,  Revion VIII
                                                Denver, Colorado
                                            STEPHEN F. TARLTON
                                                    CH2M Hill
                                                Denver, Colorado
INTRODUCTION

  The isolation of the element radium by Pierre and Marie Curie
in 1898 introduced the world to the "miracles" of radioactivity.
Pitchblende,  the mineral  from which radium was  originally  re-
fined, was known to exist in only a few locations and, consequent-
ly, the price of the material increased with the expanding use of
radium for research and treatment of cancer.
  Soon carnotite, another element with high radium content, was
discovered in the Colorado Rockies. During the early 1900s, large
amounts of carnotite ore were shipped to Europe to be refined. But
by 1912, war in Europe threatened those refining operations and,
faced with the possible shortage of refined radium, the United
States Bureau of Mines entered into an agreement with American
medical research institutions to form the National Radium Insti-
tute for the purpose of demonstrating the feasibility of radium  re-
fining and to produce enough  radium for research and medical
needs.
  The National Radium Institute conducted pilot studies  on
radium refining near the outskirts of Denver, and by 1914 a full-
scale production facility was completed and in operation. By 1918,
over 7g of pure radium had been refined and the experiment was
judged a success. The Institute was closed and the equipment sold.
Attracted by the high prices offered for radium, many other  re-
fining operations began in the  Denver area, utilizing the proven
Bureau of Mines process as well as a variety of other processes.
  The discovery of very  rich  near surface pitchblende  deposits
in South Africa proved fatal for the flourishing U.S. radium indus-
try as prices plummeted and operations ceased or were converted
to production of other materials by the mid-1920s.
  As mining went through subsequent cycles of prosperity and de-
pression, Denver grew to become the largest city in  the Rocky
Mountains, developing a diverse economic base including agricul-
ture, petroleum and the largest concentration of federal employees
outside of Washington, D.C. The population grew and the sur-
rounding prairies first became suburbs and then part of the expand-
ing metropolitan core. Defunct radium processing operations, once
located in fringe or industrial areas, were engulfed by the expand-
ing urban development. Since radium processing residues made ex-
cellent fill, they were incorporated into streets, railroad embank-
ments and other properties.
  In  1979, Hendricks of the USEPA discovered a reference to the
National Radium Institute in a  1916 U.S. Bureau of Mines docu-
ment. Subsequent investigation identified the property, and radia-
tion surveys confirmed extensive low-level radioactive contamina-
tion of the property, including buildings and soils on the site.
  Following notice of the potential for problems on other proper-
ties, the State of Colorado, the USEPA and the U.S. Department
of Energy (DOE) began a series of studies to locate the contami-
nated areas and identify the associated hazards. In 1981, the Den-
ver Radium Site, consisting of 31 identified properties, was placed
on the Superfund National Priorities List. Remedial investigations
were performed for most of the sites by the Colorado Department
of Health in 1981 and 1982. In December 1983, CH2M HILL, the
USEPA Zone II REM/FIT contractor, was directed to complete
the remedial  investigations and feasibility studies for the proper-
ties in accordance with the National Contingency Plan  (NCP) re-
quirements.
  The NCP sets forth the procedures for evaluating and selecting
remedial actions at uncontrolled hazardous waste sites. The NCP
specifies the steps to be taken as preliminary assessment, site in-
spection,  remedial investigation, feasibility studies and  design
which is followed by the remedial action.
  The feasibility study develops and considers remedial actions.
The NCP defines remedial actions as:
"Those actions consistent with permanent remedy taken in-
stead of, or in addition to, removal action in the event  of a
release or threatened release of a hazardous substance  into
the environment, to prevent or minimize the release of haz-
ardous substances so that they do not migrate to cause sub-
stantial danger to present or future public health or welfare
of the environment." (40 CFR  300.68)
  For the Denver Radium Site, the above paragraph is interpreted
to mean actions which would reduce the exposure to ionizing rad-
iation and/or radon gas from  radium  residues to prevent or min-
imize the associated danger to public health, welfare or the environ-
ment both at present and in the future.

DENVER RADIUM SITE

  The properties under study range from public streets and an alley
to parks,  vacant lots, industrial and commercial properties and  a
private residence. The actual number  of  contaminated properties
included in the Denver Radium Site has varied slightly with further
studies  and remedial actions.  Of the 31 included on the original
listing, three have been decontaminated by the owners and checked
by the Colorado Department  of Health. However, several addi-
tional properties contiguous to some  of  the original Denver Ra-
dium Site properties have been  identified for further investigation as
potentially contaminated properties because the designated proper-
ties they abut are contaminated up to, and possibly across, proper-
ty lines.
                                                                                             CASE HISTORIES
                                                       445

-------
   These continuous locations and the remaining 28 properties of
 the original 31 were grouped, the groupings based on similar char-
 acteristics  or contiguous locations.  A brief description  of the
 groupings and properties is presented below:
 Group I: Four adjacent properties which occupy a full city block in
 an industrial district. The area is used for light industrial and com-
 mercial activities including a tombstone manufacturer, equipment
 manufacturer and miscellaneous warehouse and  office space. The
 probable source of contamination for all the properties was a ra-
 dium mill operated in the early 1920s in what had been a brewery. A
 maximum value of 452 pCi/g has been measured at this site.
 Group II: Two nearly adjacent properties and surrounding areas in
 an industrial district. One operated as an analytical/research lab-
 oratory  until recently; the other is a scrap metal recycling  opera-
 tion covering a  full block and littered with massive piles of scrap
 metal. Previously, the latter property had  been  the location of a
 Radium Company of Colorado mill. The  surrounding properties
 include commercial space,  railroad tracks,  a  highway department
 equipment yard and offices. Maximum concentration of radium at
 this site was 931  pCi/g.
 Group III: Two adjacent properties  and surrounding areas in a
 mixed residential, commercial and  light-industrial neighborhood.
 On the larger property, a mill and refining operation existed until
 it was totally destroyed by fire in the  1950s. The debris remaining
 after the fire was bulldozed into the building basement and the site
 leveled.  The building across the street, reported to have been the
 mill office, is currently used by  a commercial lighting manufac-
 turer. Surrounding areas include railroad tracks, a packaging oper-
 ation and commercial  space. A value of 836  pCi/g was measured
 in soils in open  areas,  and 1956 pCi/g was measured in the crawl
 space of the building.
 Group IV: One property  in  an industrialized  area.  This  brick
 manufacturing plant occupies the site  of the original National Ra-
 dium Institute. Though most of the Institute's buildings were re-
 moved prior to the construction  of the brick plant in the  1950s,
 a small office building  and a laboratory building remain. Attached
 alpha of 632,000 dpm/100cm1 was measured in the office building.
 Group V: The railroad property contiguous to the Group IV prop-
 erty.  The properties in Groups IV and V, though physically adja-
 cent to one another, were separated for reasons of scheduling.
 Group VI: Seven open  land areas scattered throughout the Denver
 Area.  Properties include a mined gravel pit used for a landfill, a
 city park, an alley, railroad tracks, a waterline right-of-way used as
 a parking lot, a vacant lot and the grounds of a chemical  plant.
 The contaminated portion of each property is on open land, and no
 buildings or contiguous properties were identified as contaminated.
 A maximum value of 2775 pCi/g was measured on these properties.
 Group  VII: Eight street segments located  in three clusters.  The
 streets are located in urban areas near downtown Denver, passing
 through residential, and commercial areas and public parks includ-
 ing the Denver Botanical Gardens. Contamination is thought to be
 contained in a .13 cm asphalt layer in the roadways. Approximately
 38 blocks of street have elevated radioactivity, with a maximum
 reading of 51 microR/hr.
 Group VIII: A chemical plant and adjacent  railroad  properties.
 This plant has a current Radioactive  Materials License from the
 State  of Colorado and has been  in continuous use since the early
 1900s when radium was refined at  the site. Maximum values of
 0.145 WL and  2408 pCi/g were recorded  in  buildings and open
 areas, respectively.
 Group IX: A pancake restaurant, surrounding parking lot and ad-
jacent building. The site, which originally housed a radium  labor-
atory, is on one of Denver's busiest streets and is surrounded by
commercial property and residences.
Group X: A large manufacturing facility which had originally been
a cotton mill, was converted  to a radium mill and more recently
has been a mining equipment manufacturing plant.
Group XI: Originally a private residence, now used as a real estate
office. The owner of this property has done some cleanup  of the
building; however, the grounds of the property are also contam-
                                                        inated, and further remedial work may be required. A maximum
                                                        value of 199 pCi/g was originally measured at this site.
                                                        REMEDIAL ALTERNATIVES AND ISSUES
                                                          The diversity  of properties and degree of radioactive contam-
                                                        ination has led to the consideration of a variety of alternative re-
                                                        medial actions. These actions, summarized in Table 1, raise a va-
                                                        riety of issues which  complicate the evaluation and comparison of
                                                        alternatives.  Some of these issues are  briefly discussed  in  the
                                                        following paragraphs.
                                                                                    Tiblcl
                                                                   Denver Radium Sites Poulbk Remedial Actions
                                                          The actions listed below represent alternatives for cleanup of the Denver
                                                        Radium Sites thai could be considered in the preliminary screening step of
                                                        the Feasibility Study.
                                                        Contamination Removal Option*
                                                        •Material excavation and removal
                                                        •Structure demolition and removal
                                                        •Physical cleaning of surfaces (decontamination)
                                                        Disposal Options (required for removal options)
                                                        •On-site (temporary or long-term)
                                                        •Off-site (federal or commercial disposal sites)
                                                        •Reprocessing to extract uranium and radium
                                                        Radiation Shielding (Including radon control)
                                                        •Lead shielding
                                                        •Concrete mixtures
                                                        •Asphalt
                                                        •Soil
                                                        Other Actions
                                                        •Area exclusion (fencing, condemnation, and use restrictions)
                                                        •Dilution (with clean soil)
                                                        •Ventilation of enclosed areas (for radon control)
                                                        •Sealants (for radon exclusion from enclosed areas)
                                                        •No action
                                                        Combinations of (he Above Actions
                                                        •(Such  as shielding and ventilation in basements, with physical cleaning
                                                         of floors and walls)
                                                        Cleanup Standards
                                                          Since the radiation exposure from radium processing wastes has
                                                        potential health effects similar to the potential effects of exposure
                                                        to uranium mill tailings and the causal mechanisms are the same,
                                                        the USEPA has chosen to apply the inactive uranium mill tailings
                                                        cleanup standards (40 CFR Part 192) to the radium  processing
                                                        wastes present at the  Denvei Radium Site. These regulations clear-
                                                        ly emphasize remedial actions which remove the source of contam-
                                                        ination from the public; however, the NCP criteria require careful
                                                        consideration of all alternatives.
                                                          The standards in 40 CFR Part 192 provide direction for cleanup
                                                        of both structures and open land areas as follows:
                                                        •For contaminated buildings, reduce the radon decay product con-
                                                         centration to less than 0.02 Working Level and reduce the gamma
                                                         radiation to less than 20 microR/hr above background
                                                        •For contaminated open areas, reduce the surface radium content
                                                         to less than 5 pCi/g above background averaged over the top 15
                                                         cm of soil or to less than 15  pCi/g above background averaged
                                                         over any 15 cm of soil below the top 15cm.
                                                          In addition, the feasibility studies will  address compliance with
                                                        the State of Colorado Standard for maximum alpha activity from
                                                        nonuranium radioactive material. The Colorado standard is 300
                                                        dpm/100 cm1 attached  alpha and 20  dpm/100 cm' removeable
                                                        alpha.

                                                        Defining Action Levels

                                                          Applying the above standards to a given site presents several dif-
                                                        ficulties, including the selection of measurement techniques and the
 446
CASE HISTORIES

-------
determination of background levels. Devices commonly used for
field radioactivity measurements are often calibrated in units of
counts per minute or microR/hr. Conversion between these units
or to radium concentrations in soils is difficult. Thus, depending
on the field instrument used, the instrument readings may not be
suitable for comparison to the standards. Field measurement tech-
niques for determining the radium concentrations in soil are very
cumbersome, and  laboratory techniques  usually require  weeks.
Thus, during the excavation process, it is difficult to determine
whether all the contaminated material has been removed.
  Variability in  background measurements can also complicate
data evaluation.  Natural background is affected by elevation  and
the presence of naturally radioactive materials.  Background in
Denver is approximately 50% higher than at sea level due to higher
levels of cosmic radiation at its 5000 ft elevation. Natural back-
ground levels can be further elevated due to the naturally higher
uranium content of some local rock.
  At certain sites, background is more difficult to evaluate. A brick
manufacturing site is totally covered with raw clay, much of which
has a small, but measurable  naturally elevated level of radioactiv-
ity. A second site which manufactures tombstones is covered with
15 years' accumulation of granite chips,  which  also has  an  ele-
vated natural radioactivity. At this site, background readings could
also be influenced by a commercial sandblasting abrasive which
was found to contain significant radiactivity.
   Radon gas levels can also  vary due to factors that are not asso-
ciated with radioactive wastes. Buildings constructed of naturally
radioactive substances (granite, brick) or  with effective weather-
sealing may have elevated radon levels totally separate from  any
contamination-related radioactive materials.
Nonradioactive Contamination

   As illustrated above, the Denver Radium Site properties repre-
sent a cross-section of urban commercial and  industrial uses. Al-
though the initial focus of the investigation primarily involved the
problem of radioactive contamination, information about the  past
uses of some of the properties indicated that the response planning
for some of the sites should not be limited to the hazards of ra-
dium processing wastes alone. Some of the sites had been the loca-
tion of chemical processing or manufacturing operations for nearly
a century, while others had known or suspected past uses  such as
landfill or dump sites, oil recycling plants or wood preserving oper-
ations. Therefore,  for health and safety reasons  and disposal site
selection and design  purposes, investigations of  nonradioactive
hazardous materials were added to the scope of the study.
   The nonradiological hazardous materials potential of each of the
properties was  evaluated  by conducting  preliminary assessments
and, as warranted, site inspections. Although not all of the prelim-
inary assessments and site inspections have been completed, the in-
vestigations to date have revealed significant nonradioactive waste
contamination, including locally high (up to 30  ppm) concentra-
tions of carcinogenic and suspected carcinogenic  organic com-
pounds. The discovery of these organics triggered additional site
evaluation to determine the  extent of contamination and address
these substances along with the radium wastes in  the health  and
safety plan for site cleanup and in the selection and design of an
off-site disposal facility.

Disposal Site Requirements

   CERCLA required that the State assure the USEPA that a dis-
posal site is available. For the waste from the Denver Radium Site,
finding a suitable disposal site has proven to  be a difficult task.
Few hazardous  waste sites will accept radioactive wastes and haul
distances and the cost of using the commercial disposal facility at
Hanford, Washington is prohibitive.
   The similarity  between the  Denver  Radium Site radioactive
wastes and uranium mill tailings had led  to consideration of co-
disposal with uranium mill wastes at one or more locations in Col-
orado. Colorado has nine inactive uranium mill  sites under study
that are included in the DOE Uranium Mill Tailing Remedial Ac-
tion Program (UMTRAP). However, differences in the timing of
the UMTRAP cleanup work and concerns about mixed (nonradio-
active) wastes may complicate the use of an UMTRAP site. There-
fore, additional studies are  being performed to consider  suitable
alternative disposal sites.

Liability for Cleanup
  The  assignments of liability and enforcement are important as-
pects of all  Superfund actions. It is important to investigate the
existence of potentially  responsible parties that are financially vi-
able and, hence, could be compelled to cleanup a site or reimburse
the Superfund for the costs of cleanup. Several aspects of this site
involve complex and difficult responsible party issues. The first,
most obvious challenge  is tracing a chain of liability that extends
almost three quarters of a century into the past. Business and gov-
ernment record-keeping was not as complete in the past as it is to-
day. In addition, many of the records that did exist have been lost
or destroyed. In addition, any pursuit of responsible parties must
also address the difficult question of assessing liability against a
party for activities that  took place at a time when there was very
little, if any, appreciation for the health hazards associated with ex-
posure to radioactive materials.
  Investigations of liability  for the release must also address the
responsibilities of various government entities. Since the U.S. Bur-
eau of Mines was instrumental in the creation of the National Ra-
dium Institute, there is a possibility that the U.S. Government could
have some  responsibility for the release  of radium  processing
wastes. In addition to the Bureau of Mines involvement, there have
been some allegations that one of the Denver Radium Site  prop-
erties  was the location  of a thorium processing operation from
the Manhattan Project, the United States' World War II atomic
weapons program. Thus, because of possible Government involve-
ment, factual evidence related to the  liabilities of the Federal Gov-
ernment is also being collected.
  Some evidence also points  toward a liability of the City and
County of Denver for certain portions of the Denver Radium Site
releases. The consequence of such a liability is that CERCLA re-
quires  a state cost share of a  higher percentage of cleanup costs
for publicly owned sites rather than the 10% required for  other
sites. The most obvious  area of potential liability is in the contam-
inated  streets and alleys. Since they have been the property and
responsibility of the City and County of Denver since the days of
the National Radium Institute, local government or its contractors
could be held liable for any cleanup costs associated with the streets
and alleys. Local government could also face certain liabilities due
to the  spread of radium contaminated soils to otherwise uncon-
taminated fill areas owned by private parties.

Interim Measures

  Given the lead  time  required  to  complete  remedial investiga-
tions,  determine the  cost-effective  remedy  for each situation,
arrange for off-site disposal  and prepare for the cleanup,  the
USEPA has initiated  studies under its Superfund Immediate Re-
moval authority to determine whether prompt actions are needed to
protect people from exposure at  any of the Denver Radium Site
properties. This process began during the summer of 1984 when a
Technical Assistance Team  contractor investigated indoor radon
and airborne alpha particle  levels in occupied structures to deter-
mine whether temporary remedies,  such as alpha particle  fixa-
tion, installation of air cleaning devices or additional ventilation
should  be considered while arrangements are being made  for a
permanent remedy at the site.
  This study may result in recommendations to use Immediate Re-
moval  funds to take steps toward reducing exposure Levels in any
of the buildings affected by Denver Radium Site contaminants.
Since wintertime is typically the season of greatest concern due to
building weatherproofing and lack of circulation of outside air,
the time schedule for  any Immediate Removal action will empha-
size remedies that can be implemented in time to do the most good.
                                                                                                    CASE HISTORIES
                                                           447

-------
CONCLUSIONS

  The Denver Radium Site, with its numerous and diverse con-
taminated properties, presents interesting challenges to the plan-
ners of remedial actions. The radioactive wastes are  difficult to
quantify due to the complexity  of  field investigation techniques
and the degree to which the wastes have been disturbed by normal
urban activities during the decades the wastes were forgotten or ig-
nored. In addition  to the radioactive wastes, the remedial plan-
ning for the Denver Radium Site must address the problems posed
by nonradioactive hazardous materials present in the contaminated
soils. Some of the properties involved have a history of as much as
100 years of mixed industrial uses.
  The presence of such nonradioactive hazards may have a signif-
icant bearing on the health  and safety procedures used during site
cleanup and the selection and  design of disposal sites. The diver-
sity of properties involved and the uniqueness of each property's
history also present great challenges in the investigation of poten-
tially responsible parties. Finally, because of the long lead time re-
quired to address all of the issues in a study of this nature, prompt
actions should be considered in  order to reduce the exposure of
occupants of contaminated properties.
                                                        GLOSSARY
                                                          alpha particle—A decay product made up to two protons and
                                                        two neutrons.
                                                          beta  particle—A  decay product consisting  of a high-velocity
                                                        electron, usually negative in charge.
                                                          dpm—Nuclear disintegrations per minute.
                                                          gamma ray—A decay product consisting of a  photon, a quan-
                                                        tum of electromagnetic energy having a characteristic wave length
                                                        and frequency.
                                                          pCi/g—Picocuries of radium  per gram of soil; a picocurie is a
                                                        fractional portion of a curie, a  unit of radioactivity equal to 3 x
                                                        10'° disintegrations per second.
                                                          mR/hr—One millionth of one roentgen per hour; a roentgen is
                                                        the unit of radiation exposure in air due to X- or gamma rays.
                                                          working level—A  measure of exposure to short-lived radon de-
                                                        cay products in air; the working level equals  1.3  x 10' MeV of
                                                        potential alpha energy from  any combination of radon daughters
                                                        per liter of air.
448
CASE HISTORIES

-------
        SUPERFUND PLANNING  PROCESS FOR THE  OMC
       HAZARDOUS WASTE  SITE,  WAUKEGAN,  ILLINOIS

                                             JACK E. BRAUN
                           U.S. Environmental Protection Agency,  Region V
                                              Chicago, Illinois
                                           STEWART L.  DAVIS
                                                 CH2M  HILL
                                             Portland, Oregon
INTRODUCTION

  The purpose of the authors in writing this paper was to describe
the feasibility study and planning process that was conducted by
the USEPA  to evaluate cleanup  alternatives for the polychlori-
nated biphenyl contamination in  Waukegan Harbor. The Feasi-
bility Study was prepared by CH2M HILL under a national con-
tract to assist the agency on site cleanup projects.
SITE DESCRIPTION

  The OMC site is located on the western shore of Lake Michigan
in Waukegan, Illinois.  It is about 37 miles north of Chicago and
about 10 miles south of the Wisconsin border (Fig. 1).
  For the purposes of the feasibility study, the site was divided into
foursubareas:
•Slip No. 3
•Upper Harbor
•North Ditch area
•Parking Lot
  High levels of PCBs  in soil and harbor sediments in the vicinity
of the OMC plant were  discovered in 1976. The movement of PCBs
through groundwater and surface water has contributed to Wauke-
gan Harbor  and Lake Michigan contamination; they have en-
tered the aquatic food  chain, accumulating in game and commer-
cial fish.
  The extent of cleanup is based on computer models of trans-
port mechanisms in the area and the application of the regulations
intheNCP.
  The contamination of the sediments in Slip No. 3  ranges from
500 to 10,000 ppm PCB. Concentrations in excess of 10,000 ppm
PCB occur in one localized area  near the former industrial out-
fall. Approximately 305,200 Ib of PCBs are in  10,900 yd3 of sedi-
ment in Slip No. 3. Because slip No. 3 contains very high concen-
trations  of PCBs in a  relatively small area, the engineering ap-
proach taken to reduce or eliminate the spread of PCBs is to re-
duce the contact of sediment and water and to reduce the concen-
tration gradient that forces PCBs downward into the sediment.
  The North Ditch drainage consists of  areas  of high PCB con-
centrations that are exposed to flowing surface water. The flow-
ing waters, especially during storms, can wash the PCBs into Lake
Michigan. Two subareas of the drainage also contain PCBs in high
concentrations.  This extreme concentration gradient forces PCBs
to move downward into deeper sediments and the groundwater.
  The engineering approach to reduce or eliminate the release of
PCBs would  be to eliminate their  exposure to the flowing surface
water and to reduce the concentration gradient in the areas of high-
est contamination.
  The Parking Lot area contains contamination of soils that ranges
from 50 to over 5,000 ppm PCBs. Approximately 277,700 Ib of
PCBs are contained in 105,800 yd3 of soil.
  The Parking Lot contains lower concentrations of PCBs distrib-
uted  in  a very large volume of soil and over a large area. In the
Parking Lot area, the pathway for the release of PCBs is through
groundwater movement toward the lake.
  The engineering approach to reduce or eliminate PCB  spread
from the Parking Lot would be to isolate the groundwater from
contacting the contaminated material.
              •••- ; I. I I           || (—I NORTH SHORE
            North Ditch Area -^Jj Q 5S?ic7    0
                            	
                          Upper Waukegan
                          Harbor Area
                        Figure 1
                        Site Map
                                                                                        CASE HISTORIES
                                                    449

-------
          o  In-Place Destruction
          o  In-Place Fixation
          o  In-Place Separation and Removal
          o  Dredging
          o  Excavation
          o  Sediment Dispersal Control
          o  Surface and Ground Water Control
          o  Bypass
          o  Dewatering
          o  Fixation
          o  Water Treatment
          o  Onsite Storage/Disposal
          o  Offsite Storage/Disposal
          o  Transportation
                          Figure 2
              Processes Evaluated in Feasibility Study

  The Upper Harbor  contamination  of sediments  with  PCBs
ranges from 50 to 500 ppm. Approximately 5,000 Ib of PCBs are
distributed in 35,700 yd1 of sediment.
  The Upper Harbor contains lower concentrations of PCBs dis-
tributed in a very large volume of sediment and over a much greater
surface area. This increased surface area provides  a significant
pathway whereby PCBs can contact the water and move out into
Lake Michigan or enter the food chain.
  The engineering approach to reduce  or eliminate the spread of
PCBs  from  the Upper Harbor is  to reduce the surface area  by
which  PCBs contact the water. The large volume, low concentra-
tions and large surface area are considered.
  The North Ditch area contamination of soils ranges from  50 to
10,000 ppm  PCBs. Concentrations exceeding 10,000 ppm occur in
one  localized area.  Approximately 495,500 Ib of PCBs are  in
70,800 yd' of soil in the North Ditch area.
EVALUATION PROCESS
  The purpose of the feasibility study was to review possible ways
of treating the problem, to screen alternative actions and to present
recommended alternatives for public review and comment. Over 70
possible processes were initially evaluated to determine their poten-
tial for contributing to PCB removal. These included various com-
binations of the technologies shown in Figure 2.
  The processes  retained from the preliminary screening  were
assembled into various combinations resulting in 21 cleanup  al-
ternatives for further study. How various processes were combined
into one cleanup alternative for the North  Ditch area is shown in
Figure 3.
  Finally, 12 alternatives and two subalternatives were selected for
a more detailed examination. All of these alternatives  were eval-
uated on the basis of the overall project objectives including:
•Effectiveness in cleaning up the site
•Technological feasibility
•Environmental soundness
•Protection of public health and welfare
•Cost-effectiveness
SUMMARY OF FINAL ALTERNATIVES
  In Slip No. 3, the reviewed alternatives included excavating deep
sediment, followed by dewatering in lagoons or barges, solidifica-
tion of the sediment and transport off-site for disposal. These al-
ternatives range from $7.6 million to $10.7 million.
                                                        In the Upper Harbor,  similar alternatives were reviewed. Be-
                                                      cause of the greatly increased volume of sediment, costs ranged
                                                      from $13.6 million to $24.8 million.
                                                        Alternatives were also developed to address the Upper Harbor
                                                      and  Slip No. 3 together.  These alternatives included a combina-
                                                      tion  of off-site removal and on-site containment. The costs ranged
                                                      from $6.1 million to $9.3 million. Removal and off-site disposal of
                                                      the most highly contaminated material from Slip No. 3 was consid-
                                                      ered  as a subalternative, increasing the costs to $12.4 million.
                                                        Alternatives addressing the North Ditch and Parking Lot area in-
                                                      cluded combinations of excavation, off-site disposal, on-site con-
                                                      tainment,  capping and construction of a storm bypass sewer.
                                                      These alternatives ranged from $740,000 for hot spot removal in
                                                      the North Ditch area alone to $62 million for the North Ditch and
                                                      Parking Lot areas together.
                                                        The combination of alternatives that was initially suggested to
                                                      the public for implementation  is shown in Figure 4.  Under this
                                                      recommendation, "hot spot" removal was advocated for Slip No.
                                                      3 and the North Ditch area. Dredging was advised for the Upper
                                                      Harbor with dredge material being permanently placed in Slip No.
                                                      3. This  process would have resulted in part of Slip No. 3 being
                                                      permanently filled  in.  A  storm  sewer bypass was recommended
                                                      for the North Ditch, and in-place containment was recommended
                                                      for the Parking Lot area.
                                                      COMMUNITY RELATIONS PROGRAM
                                                        The USEPA conducted an extensive community relations pro-
                                                      gram, allowing the  public an opportunity to comment on the pro-
                                                      posed plan.  Following the first public  comment  period, and in
                                                      conjunction  with some additional information, the USEPA en-
                                                      larged  the recommended  scope  for  harbor dredging, suggesting
                                                      permanent encapsulation in  Slip No. 3. The larger dredging pro-
                                                      ject necessitated a larger containment cell. This option was again
                                                      presented to the public for review. The Agency recieved significant
                                                                                Figure 3
                                                                            North Ditch Area
                                                                            Alternative No. 4
                                                        Cleanup  Action
Estimated Cost
                                                        SLIP NO. 3 AND UPPER HARBOR
                                                         Remove Hoi Spoil and Dispose  OHsile	$3,150,000
                                                        SLIP NO. 3 AND UPPER HARBOR
                                                         Dredge, Dewater, and Dispose In Slip No. 3 	  6,100,000
                                                        NORTH DITCH AREA
                                                         Remove Hot Spots and Dispose  OHsite	   740,000
                                                        NORTH DITCH AREA
                                                         Contain and Cap	  4,210,000
                                                        PARKING LOT
                                                         Contain and Cap	  3,210,000

                                                        TOTAL ESTIMATED COST FOR
                                                        THE FIVE  CLEANUP ACTIONS	
     $17,410,000
                                                                                Figure 4
                                                                             Estimated Costs
                                                                   Alternative Selected for Implementation
450
CASE HISTORIES

-------
objection to this  larger containment  cell during the comment
period. The USEPA then tried to maintain the enlarged scope of
the project requested during the first comment period, but reduce
impacts on the harbor, which was the thrust of the comments dur-
ing the second comment period, while simultaneously maintaining
the cost-effectiveness of the project. This effort required prepara-
tion of an addendum to the feasibility project.
  Also, during the decisionmaking process, the USEPA evaluated
the "fund balancing"  criteria of the NCP. Under these criteria,
the USEPA can recommend a less costly alternative that provides
similar protection  if funds spent at one site would preclude re-
sponse actions at other equally significant areas.
  Therefore, on this project, the USEPA identified a cost-effective
alternative and decided to implement a fund-balanced approach.


PLAN SELECTED FOR IMPLEMENTATION

  First, in Slip No. 3,  5,700 yd3 of the most highly contaminated
sediment from the localized "hot spot" will  be dredged for off-
site disposal. The estimated cost will be $3,150,000.
  Next, the remainder  of Slip No. 3 and the upper harbor will be
dredged. All sediment exceeding 50 ppm PCB will be dredged and
removed. A dewatering lagoon will be built on the OMC property
adjacent to the harbor.  The sediment  will be dewatered in the
lagoon and transported to the Parking Lot area for codisposal.
This will cost an estimated $10 million.
  In the North Ditch, hot spot removal and off-site disposal are
recommended  for  implementation,  with  5,500  yd3 of the  most
highly contaminated material to be excavated and hauled off-site
for disposal in a chemical waste disposal site. It is estimated that
this portion of the work will cost $740,000.
  Next, slurry walls will be constructed to create an on-site contain-
ment cell in the vicinity of the hot  spot removal. Contaminated
sediment will be excavated for sewer construct. This sediment will
be placed in the containment cell and will be capped in place. Then,
a storm bypass sewer will be installed in the North Ditch. This con-
struction will cost an estimated $4,210,000.
  Lastly, the contaminated soil currently in the  Parking Lot, the
sediment from the harbor and the contaminated lagoon material
will be disposed of together in the  Parking Lot. Slurry walls and a
clay cap will be used to contain the contaminated material. The
estimated cost for this portion of the project will be $3,200,000.
  The total estimated project costs  will be $21,300,000.
  Combined, these projects  will remove more than 92% of the
PCBs from Slip No. 3 and the Upper Harbor. They will also re-
move 57% of the PCBs in the North Ditch and Parking Lot area
for off-site disposal. More than 98%  of all the PCBs on-site will be
effectively isolated from contact with the environment.
                                                                                                 CASE HISTORIES
                                                         451

-------
        SITE  CONDITIONS  AND  CORRECTIVE ACTION AT
                       THE  NORTH  HOLLYWOOD  DUMP

                                       JOHN  D. TEWHEY, Ph.D.
                                              E.G. Jordan Co.
                                              Portland, Maine
                                         ANDREW F. McCLURE
                                             NUS Corporation
                                          Pittsburgh, Pennsylvania
                                          TERRY K. COTHRON
                           Tennessee Department  of Health and Environment
                                            Nashville, Tennessee
INTRODUCTION

  The North Hollywood Dump is located on the alluvial plain and
in abandoned channels of the Wolf River in Memphis, Tennessee
(Fig. 1). An uncontrolled  dumping operation existed in the early
years (mid-1930s onward), but aerial photographs of the site in the
1950s and 1960s show controlled access and a managed dumping
operation. Now closed, the dump occupies over 70 acres, and the
r~,&Łi NOflTH HOLLYWOOD DUMP SITE
                                SUBFACE WATER BOOIi'j
           HIS, TCNN UM9)
                        Figure 1
 Site Location of the North Hollywood Dump in North-Central Memphis,
  Tennessee. North Hollywood Street was Built Through the Dump in the
                       Early 1960s
thickness of the refuse is estimated to be 20 to 25 ft. Most of the
dump is covered with grasses, brush and trees. Approximately 5%
of the dump area is not vegetated due to  the presence of demo-
lition debris, tar and other materials that will not support vegeta-
tion.
  The USEPA conducted studies at the North Hollywood Dump
and in the vicinity in 1979 and 1980. On the basis of their findings,
a voluntary Technical Action Group (TAG) composed of the City
of Memphis, Velsicol  Chemical  Corporation,  Memphis/Shelby
County Health Department and the State of Tennessee joined with
the USEPA in 1980 to address problems at the site.
  Short-term assessment of the dump was addressed first. A de-
tailed site survey on a grid basis was undertaken, and a work plan
for securing surface contaminants was developed.'  In  February,
1981, 89 drums of contaminated surface soils from the eastern por-
tion of the dump were shipped to an approved landfill for disposal.
The reseeding of newly-covered portions of the dump  was com-
pleted in March, 1981.
  The long-term objectives of the TAG  Environmental Assess-
ment and Action Plan for the North Hollywood Dump were estab-
lished in  March, 1981.  The objectives are  direct and concise, yet
comprehensive. The goals were to:
•Determine the nature and scope of environmental problems exist-
 ing on the dump and adjacent areas
•Determine the relationship of the dump to  these problems
•Develop alternatives to remediate identified problems attributable
 to the dump
  To accomplish the objectives, the investigative phase of the TAG
program was divided into six major Task  Elements:
•Existing Data Compilation and Interpretation
•Surface and Subsurface Hydrology
•Subsurface Water Quality
•Environmental Monitoring
•Report Preparation
•Program Management
  TAG participants were involved in data  collection and program
management. The  E.C. Jordan Co. was subcontracted by the
USEPA's contractor,  NUS Corporation,  to undertake the fifth
task  report preparation.2 '  The company was hired to interpret
site data and identify and evaluate corrective action alternatives.

PHYSICAL SKTTING

  The dump is bordered on the north  by the westward-flowing
Wolf River, on the east and west by surface water ponds and on the
south by a residential area. The subsurface geology  in the vicinity
452      CASE HISTORIES

-------
                      NORTH
                                                                                                      SOUTH
                           MONITORING
                                        NORTH
                                      HOLLYWOOD
                                        DUMP
                                                                  TERRACE DEPOSITS
                                                                   JACKSON FORMATION
                                                                             1000
                                                            Figure 2
                                North-South Cross Section Through the North Hollywood Dump and Vicinity
 of the dump, from youngest strata to oldest, consists of the follow-
 ing (Fig. 2):
 •Alluvium—surficial sand and gravel deposits
 •Loess—windblown silt deposits which are present over most of the
  Memphis area
 •Terrace deposits—sand and gravel deposited by ancestrial streams
  and rivers
 •Jackson Formation—200 ft thick deposit of marine clay
 •Memphis Sand—700  ft thick sandy formation which serves as the
  principal aquifer for the city of Memphis.
^.;,	•-   '     $  /"I,   «|
?L.'_ ••••   -     <Ł>  /    '    !
                                         |:'^.1*T"~   Pumpin;
                                            '''•""•j:    HOLL
                                       GROUNOWATER LEVEL MEASUREMENTS
                                       MADE ON JUNE 17. 1983
 — 210— EQUIPOTENTIAL LINE

 	^GHOUNDWATER FLOW LINE

   •0-  MONITORING WELL LOCATIONS


                          Figure 3
 Horizontal Groundwater Flow Net of the Shallow Groundwater Regime in
                 Alluvium and Terrace Deposits
  The principal surface water bodies in the vicinity of the dump are
the Wolf River; an abandoned dredge pond (40 acres), beaver pond
(0.4 acres)  and oxbow lake (2 acres) on  the east; and an active
dredge pond (32 acres) on the  west. Surface water courses emanat-
ing from the dump flow to these adjacent water bodies.
  At least two groundwater regimes exist beneath the North Holly-
wood Dump:  (1) a shallow aquifer in alluvium and terrace deposits
(Fig. 3), and (2) a deep aquifer in the Memphis Sand. Groundwater
in the shallow aquifer flows to  the north beneath the dump and dis-
charges to  the Wolf River and adjacent water bodies. The deep
aquifer is isolated from the shallow aquifer by the thick clay de-
posits  of the Jackson Formation. Therefore, any contaminated
groundwater in the shallow regime which might emanate from the
dump does  not migrate to the deep aquifer in the Memphis Sand.
  The  TAG  have used  monitoring studies to compile chemical
data related to the North Hollywood Dump. Low to moderate lev-
els of chlorinated cyclodienes  have been detected in some surface
soil, surface water, sediment, groundwater and fish samples.

ENVIRONMENTAL PROBLEMS

  The  quantification and interrelationships of several factors were
considered  when  determining the nature and scope of environ-
mental problems existing at the dump and in adjacent areas. These
factors included:  (1) chemical transport  modes, routes and rates
(Fig. 4); (2) absolute concentrations and  concentration gradients
of chemicals  in the environment; (3) anticipated duration of ex-
posure; (4) potential receptors; and (5) health and environmental
quality risks.  On the basis of the assessment of these five factors,
the following environmental problems were ientified:
•On-Site Soils—residual surficial contamination may pose a risk to
 public health and environmental quality; additionally, contami-
 nants  will  be a future threat to adjacent surface water bodies via
 surface runoff.
•Surface Water Bodies—the presence of chlorinated cyclodienes
 in adjacent surface water ponds (maximum measured concentra-
 tion of 0.001 mg/1, in associated bottom sediments  (maximum
 concentration of 200 mg/1) and  in whole fish (maximum concen-
 tration of  28 ppm) constitutes a potential health risk.
•Shallow  Groundwater—the  contaminants present within the
 dump could result in increased  contaminant transport to  adja-
 cent surface water bodies via shallow groundwater flow and dis-
 charge.

REMEDIAL RESPONSE OBJECTIVES

  The  identification and evaluation of corrective action  alterna-
tives for existing or potential environmental problems at the North
Hollywood Dump requires the establishment of appropriate remed-
ial response objectives and cleanup goals. Chlorinated cyclodienes
                                                                                                  CASE HISTORIES
                                                         453

-------
                                                                                  ALLUVIUM/TERRACE DEPOSITS
                                                             Figure 4
Conceptual Pathways for Contaminant Migration at the North Hollywood Dump. Environmental Problems Have Been Found to be Associated with SurfitiaJ
  Soils on the Dump Itself and Surface Water, Bottom Sediment and Fish in Adjacenl Ponds. Shallow Groundwater Emanating from the Site is a Potential
                                                      Environmental Problem.
are the key chemicals of concern since this class of organic com-
pounds includes many pesticides and related compounds, generally
exhibits a low solubility in water, has a high affinity for soil, has a
low volatility and undergoes natural degradation.
   Metals have been detected in the soils, sediments and waters at
the site. However, their concentrations and environmental impact
are considered less  significant than those of the chlorinated cyclo-
dienes. Moreover, remedial actions that mitigate the environmen-
tal effects  of  the chlorinated  cyclodienes can also  be expected to
mitigate effects of inorganic substances.


ON-SITE SOILS
   Very few specific, quantitative health criteria exist for the assess-
ment of potential exposure routes via contact and/or ingestion of
contaminated  surface soil. Several independent criteria in the form
of guidelines,  experimental  data and background studies relating
to health risk  via soil contact  include: (1) Work Health Organiza-
tion  Guidelines for limits  of pesticide ingestion based on body
weight,4  (2) lethal dosage (LD) experiments on animals' and (3)
background levels of pesticides and related compounds in the en-
vironment.5'6
  Concentrations of total chlorinated cyclodienes in the soil range
from less than 1 ppm to as high as 100 ppm. The lower value is close
to background levels measured in several U.S. cities,'16 while the
higher value (100 ppm) is an order of magnitude higher than the
highest background levels in U.S. cities.5'6 Jordan considered  it
appropriate to identify a narrow range of total concentration values
as a response objective (10 ppm to 100 ppm) rather than a specific
value within the range. A specific response objective within the
range of values would  be established on the basis  of anticipated
future land use options represented as "no use," "limited use" and
"unrestricted use". A system response curve for surface soil under
a "no action" scenario is shown in Figure 5.
                                                                                                      ' ARROW DENOTES TEW I
                                                                                                      , f WHICH SYSTEM MEETS
                                                                                                        RESPONSE OBJECTIVE •
                                                                                                        NATURAL OEORADATIOII
                                                                   IIItfMO COMOlTIOMI
                                                                                           OCJCCTIVt
                                                                                           'ACf *O«.t
                                                                                          I10O •»»!
                                                                                      CALENDAR YEAR
                                                                                     Figure 5
                                                          The System Response Curve for the North Hollywood Dump Surficial Soil
                                                             Indicates the Length of Time Necessary for Response Objectives of
                                                             10 ppm and 100 ppm Total Chlorinated Cyclodienes to be Met Via
                                                                   Natural Degradation Assuming a 10-Year Half Life.
454
CASE HISTORIES

-------
Adjacent Surface Waters

  Most chlorinated cyclodienes in surface water bodies are in the
bottom sediments and are likely in equilibrium with the overlying
water. Jordan calculated that the distribution ratio of total chlor-
inated cyclodienes between sediments and  surface waters  in the
large abandoned dredge  pond located to the  east of the dump is
approximately 100,000.
  Fish in the surface waters adjacent to the dump derive contam-
inants from water, sediment and aquatic  biota, including other
fish. The relative  contribution from each source, and  therefore
the relative level of contamination, is dependent on the species of
fish and is largely controlled by feeding habits.
  Water quality criteria define the  requisite quality of surface
water,  and  USFDA  action levels specify  allowable  fish  levels.
USFDA action levels for edible portions of fish and water quality
criteria that have been established for the protection of freshwater
aquatic biota are not independent entities. The water quality cri-
terion (WQC) for chlordane was derived  from  the established
USFDA action level7 according to the following relationships:
USFDA action level for edible portion
                                      =  WQC
(1)
bioconcentration factor
   lipid* in whole fish
                       x % lipids in edible portion
 The bioconcentration factor (BCF) is represented by:

 concentration of chemical in whole fish
 	 = BCF
 concentration of chemical in water
(2)
  The relationship between water quality criteria and USFDA ac-
 tion levels permits one to calculate water quality criteria for  site-
 specific conditions if sufficient data are available.  In the case of
 surface waters adjacent to the North Hollywood Dump, response
 objectives were established for surface water on the basis of water
 quality criteria (0.0043 ug/1 chlordane)7 and for fish on the basis of
 USFDA action levels (0.3 ppm chlordane).8 Chlordane is consid-
 ered to be an appropriate compound upon which to base a response
 objective because technical  grade chlordane  components consti-
 tute the majority of chlorinated cyclodienes found in most analyses
 conducted by TAG.
  Because bottom sediments represent a source of contaminants
 for water and fish, the attainment of response objectives for sur-
 face water and fish must ultimately depend on the attainment of a
 cleanup goal for  sediments. The crux then becomes... how clean
 must the bottom sediment be to attain contaminant concentrations
 that do not exceed water quality criteria levels in surface water and,
 in turn, USFDA action levels in fish?  The options for cleanup of
 the sediments in adjacent surface water bodies are:
 •Total removal or containment of an appropriate thickness of bot-
 tom sediments over the entire area of the contaminated surface
 water bodies
 •Partial removal or  containment of bottom  sediments  based on
 chemical relationships such as: (1) water  quality criteria  which
 link fish quality to water quality and  (2) partitioning coefficients
 for chemicals in coexisting media which links  water quality to sed-
 iment quality at equilibrium conditions
 •Partial removal of  contaminants based  on results of additional
 analytical data which focus on  those areas that have the greatest
 degree of contamination
*Lipids are fats or fatlike substances in fish that are capable of solubilizing and storing organic
substances such as chlorinated cyclodienes that are ingested by the fish. Different fish species
have different percentages of lipid content by weight. Lipids are distributed in the organ cavity as
well as in edible flesh. USFDA action levels reflect contaminants present in lipids that are asso-
ciated with the edible portion (flesh) of fish.
                   EXPOSURE UNITS (EU) FOR QUANTIFICATION OF
                   RISK FROM CONCENTRATIONS OF DEGRADABLE
                   MATERIALS PRESENT IN SURFACE ENVIRONMENTS
                   ARE CALCULATED ON THE BASIS OF
                    • CONCENTRATION ABOVE A REFERENCE LEVEL
                     SUCH AS THE RESPONSE OBJECTIVE
                    • AREA OVER WHICH THE RISK IS POSED

                    . ANTICIPATED TIME DURATION OF EXPOSURE
            SHADED VOLUME REPRESENTS THE QUANTITY OF EXPOSURE UNITS.
            NUMBER OF UNITS CHANGE FROM TIME A TO TIME B DUE TO
            NATURAL DEGRADATION.
NOTE'
CONC.
THE CONCENTRATION -TIME PORTION OF THE EU
PLOT REPRESENTS A SYSTEM RESPONSE CURVE
/"Tyy*^.^— 	 SYSTEM RESPONSE CURVE
//////?J*}~^ f RESPONSE OBJECTIVE
^^
TIME
                                   Figure 6
                      Explanation of "Exposure Units" (EU).

        Shallow Groundwater

          Shallow groundwater at the North Hollywood Dump represents
        a potential environmental  problem rather than an actual one.
        Hence, no cleanup goals were established for the shallow ground-
        water regime.
        CORRECTIVE ACTION ALTERNATIVES

          In the  initial screening of available remedial technologies, the
        principal  criteria used were: (1) applicability to site conditions and
        (2)  reliability/practicality.  After the development  of corrective
        action alternatives, an evaluation done of the 41 possible cleanup
        alternatives was made. The evaluation criteria utilized were:
        •Technical feasibility
        •Implementability
        •Protection of public health
        •Protection of environmental quality
        •Time to  attain response objectives
        •Duration and intensity of field monitoring requirements
        •Effect on future land use
        •Cost [capital, operations  and maintenance (O&M) and  present
         worth]

        Evaluation of Corrective Action Alternatives

          Corrective action alternatives were evaluated for  four  specific
        problem areas  associated with the North Hollywood Dump: (1) the
                                                                                                      CASE HISTORIES
                                                                    455

-------
                        5.«H
                                              ID 1IPOIUM Out TO
                                               COVIM ON DUMP AMA 4
                               /    /    ^ (xt)HAMVItT fWM (tlf.OOO)
                                             innuno ACCIM pimuiTin ROAD POX piAcma afOPA»ic«i«.o«»

                                                -cOMtTKucr von* 'ONO i
                                             KIIRIUOVI POHO 0»*U. PlACi OIOPA»IC A«D COVU (HI.OOOl
                                             100      IftO      ?OO      2(0   v   30O      350


                                            PRESENT WOOTM Of AtTEBNATIVIt IN l»3 OOLIABS (. IOOO)
                                                              Figure 7
               Example of Cost-Effectiveness Diagram for the Various Corrective Actions Identified and Evaluated for the Oxbow Lake
dump itself; (2) the 40 acre abandoned dredge pond located east of
the dump; (3) the 0.4 acre beaver pond; and (4) the 2 acre oxbow
lake. The alternatives evaluated  by Jordan for the  dump surface
and the oxbow lake are found in Table 1.  Alternatives considered
for the abandoned dredge pond and the beaver pond are similar in
principle in those for the oxbow lake.
  The mitigation of contaminants present at the site is dependent
on  two principal factors: (1) the natural degradation rate of the
chlorinated cyclodienes and (2) the quantity of contaminated ma-
terials removed  or otherwise constrained at the site. A quantita-
tive measure of the relative magnitude of the environmental prob-
lems at the site and the potential  chemical exposure was developed
by  Jordan. This measure of exposure, which is referred to herein
as "exposure units," constitutes  the common basis  for measuring
the effectiveness of various corrective action alternatives for prob-

                            Table 1
    Corrective Action Alternatives for Selected Problem Areas of tbe
                     North Hollywood Dump

North Hollywood Dump
•No action
•Fence site until contaminants in surface soils decay naturally to less than
 response objective
•Cover all surface areas in which contaminant levels are  greater than re-
 sponse objective. (Fencing option also considered)
•Cover all surface areas in watersheds on dump contributing to adjacent
 surface water ponds. (Fencing option also considered)
•Cover entire dump. (Fencing option  also considered)
•Remove and dispose of soils in which contaminant levels are greater than
 response objective
Oxbow Lake
•No action
•Fence lake until contaminants in bottom sediments decay naturally to less
 than response objective
•Harvest fish  periodically until contaminants in bottom sediments decay
 naturally to less than response objective. (Fencing option also considered)
•Dredge bottom sediments to meet response objectives
•Excavate bottom sediments to meet response objectives
•Cover bottom sediments with clean fill
•Cover bottom sediments with geofabric and clean fill
•Fill pond with clean fill
                                                          lem areas (Fig. 6). The quantity of exposure units can be calculated
                                                          for an area on the basis of:
                                                          •Concentration of contaminants above an established reference
                                                            level such as a response objective
                                                          •Area over which the risk is posed
                                                          •Anticipated time duration or persistence of exposure
                                                             The concept of "exposure units" was utilized in the assessment
                                                          of cost-effectiveness  of corrective  action  alternatives. Two  im-
                                                          portant  aspects of the  corrective action alternatives study were:
                                                          (1) the mitigation of exposure (i.e., the reduction of exposure units)
                                                          and (2) the cost of alternatives. Plots of cost against exposure units
                                                          (EU), such as shown in Figure 7,  were developed for all the en-
                                                          vironmental problem areas of the site. The advantage of the cost-
                                                          EU graphs is that both cost and effectiveness of corrective actions
                                                          are illustrated simultaneously.  Costs  are  shown as  dollars ex-
                                                          pended, and effectiveness is shown as reduction in exposure. The
                                                          time  factor, i.e., the time required to achieve a reduction  in ex-
                                                          posure, is not indicated on the graphs.

                                                          ACKNOWLEDGMENTS

                                                             The comments and suggestions of the members of the Technical
                                                          Action Group for the North Hollywood Dump contributed signifi-
                                                          cantly to the quality of the Task Element V reports prepared by E.
                                                          C. Jordan Co.

                                                          REFERENCES

                                                          1. Technical Action Group, North Hollywood Dump, Work Plan for Se-
                                                             curing Surface Contaminants at Hollywood Dump, 1981.
                                                          2. E.C. Jordan Co., Task Element V-A, Data Interpretation Report, pre-
                                                             pared for NUS Corporation, FEb., 1984.
                                                          3.  E.C. Jordan Co., Task Element V-B,  Corrective Action Alternatives
                                                             Report, prepared for NUS Corporation, Feb., 1984.
                                                          4. Edwards,  C.A., Persistent Pesticides in the Environment, Chemical
                                                              Rubber Co. (CRC) Press, Cleveland, OH, 1970,19.
                                                          5.  Wiersma,  O.B., Tai, H. and Sand, P.F., "Pesticides in Soils", Pesti-
                                                             cides Monitoring J., 6. 1972,126-129.
                                                          6.  Carey, A.E., Wiersma, G.B. and Tai,  H., "Residues in Soils", Pesti-
                                                              cides Monitoring J., 10, 1976, p. 126-129.
                                                          7.  USEPA, Ambient  Water Quality Criteria for CMordane, EPA  440/
                                                              5-80-027, Washington, DC, 1980.
                                                          8.  Crowell, R., USFDA, Personal communication, 1984.
456
CASE HISTORIES

-------
           KADON  CONTAMINATION IN MONTCLAIR  AND
              GLEN RIDGE, NEW JERSEY: INVESTIGATION
                            AND  EMERGENCY  RESPONSE

                                              JOHN V. CZAPOR
                                           KENNETH GIGLIELLO
                                   U.S. Environmental Protection Agency
                                                    Region II
                                             New York, New York
                                               JEANETTE ENG
                           New Jersey Department of Environmental Protection
                                       Bureau of Radiation Protection
                                              Trenton, New Jersey
 INTRODUCTION

  In 1979, the New Jersey Department of Environmental Protec-
 tion (NJDEP) initiated a program to identify and investigate those
 locations in the state which were once the site of radium processing
 facilities.  Chief among those sites investigated was a facility in
 Orange, New Jersey which had ceased operation in the 1920s. Con-
 cern over the possibility  of off-site disposal of processing waste
 prompted an aerial gamma radiation survey of surrounding areas
 of Essex County. This survey identified a number of areas having
 high gamma radiation; the level and extent of several of the gamma
 anomalies suggested the possibility of waste disposal. In July, 1983,
 NJDEP began to investigate two areas of high gamma readings in
 the towns of Montclair and Glen Ridge. The results of this in-
 vestigation became available in late November, 1983, and identified
 a number of homes having unacceptably high levels of radon gas.
  On Dec. 6, 1983, the  USEPA, acting under the mandates of
 CERCLA, began immediate removal actions at both the Glen
 Ridge and Montclair sites. Response actions were taken to reduce
 the residents' exposure to radon gas and radon progeny.  Concur-
 rent with the response action, the USEPA and NJDEP developed
 and implemented an extensive field investigation to totally assess
 the extent of the problem.
  While radon problems have been identified and addressed in
 several areas of the country (e.g., uranium mill tailings in  the west
 and phosphate tailings in the southeast), the potential threat posed
 by  radioactive processing  material disposed  of  in  densely
 populated, urban residential communities poses unique problems.
 In this paper, the authors discuss both the immediate removal ac-
 tions and the field investigation initiated in response to the prob-
 lems in Montclair and Glen Ridge.
 INITIAL SURVEY ACTIVITIES

  In the first quarter of  the 20th century,  an early radiation in-
 dustry existed in Essex County, New Jersey. In 1979, the  NJDEP
 began investigating a  former  radium  processing  facility (U.S.
 Radium) site in Orange, New Jersey; the investigation resulted in
 the  site's inclusion on the USEPA National Priorities  List  for
 Superfund remedial action.
  Historical records showed that this facility operated for approx-
 imately ten years (1915-1926) and processed about one-half ton of
 ore daily. The NJDEP was concerned that such a rate of  ore pro-
 cessing coupled with the limited size of the original property might
indicate the possible use of nearby non-contiguous areas  for
disposal of processing waste.  At the  request of NJDEP,  the
USEPA conducted an aerial gamma radiation survey of a 12 mi2
area surrounding the  facility. The resultant isoexposure  contour
map identified areas of elevated gamma radiation and suggested the
existence of several possible areas of waste disposal. The areas were
in the towns of Glen Ridge, Montclair and West Orange.
  In July, 1983, the NJDEP, after consultation with local officials,
began a preliminary investigation of the Montclair and Glen Ridge
sites. An initial outdoor gamma radiation survey was conducted in
publicly accessible areas. Gamma radiation levels greater than 15
jtR/hr (micro-Roentgens per hour), a level approximately twice
background,  identified properties to be investigated further;  per-
mission of owners was then obtained for further surveys of the pro-
perties. Affected properties were  surveyed  using gamma
measurements at the surface and at 3 ft above the ground surface.
Shallow boreholes were drilled using a hand augur;  the gamma
radiation in the holes was measured and soil samples were taken for
laboratory analysis. In addition, an indoor radiological survey was
conducted including  indoor gamma  measurements and  the
measurement of radon gas using activated carbon canisters.
  The  gamma  survey and subsurface coring program identified
several areas  in  both  neighborhoods  in  which contaminated
material had been disposed. The indoor survey identified a number
of houses with radon concentration levels well in excess of the ex-
pected background range. Several houses surveyed had levels which
exceeded the radon concentration equivalent for radiation workers.
The information available was sufficient to indicate that there was
an imminent and substantial endangerment to public health and to
support the initiation of a CERCLA removal action.

RISK ASSESSMENT AND
MANAGEMENT PLAN

  The risks associated with exposure to elevated levels of radon gas
and radon decay products (also known as "radon daughters" or
"radon progeny") have been extensively discussed in the literature.
Several sources of information are provided as references to this
paper. In brief, radon is a radioactive gas produced by the decay of
radium-226, a naturally occurring element. The radon  decays  into
short-lived radon progeny; these  progeny  are charged particles
which may attach to particles in the air. When inhaled and retained
in the lungs, the progeny decay further, emitting alpha radiation.
The health effect associated with this exposure is an increased risk
of lung cancer.  Radon quickly disperses  in  the ambient  air;
however, in enclosed areas without adequate ventilation, the gas
can  concentrate, resulting in levels of radon and radon progeny
posing a significant risk of increased lung cancer.
  In early December, 1983, state and federal public health and en-
vironmental officials met to develop a risk assessment and manage-
ment plan. The results of those meeting were summarized in the
                                                                                          CASE HISTORIES
                                                     457

-------
Public Health Advisory for Glen Ridgc/Montclair,  New  Jersey
issued by the Centers for Disease Control (CDC).
  Specifically, the Health Advisory developed risk estimates  for ex-
posure of the residents. Based on a number of sources, the CDC
calculated the annual risk of lung cancer for residential exposure to
selected radon levels (Table 1).
  The risk assessment and  management plan  also outlined criteria
by  which  all residences  sampled  would  be divided into  four
categories or tiers based on the extent of the radon problem in the
home; the required actions  and time frames for immediate removal
actions  for each  tier were  also  identified. The time frames were
established so that no resident was unduly exposed to a risk greater
than  1/105  during the response  period. The  tier categorization is
summarized in Table 2.
  Finally, the risk assessment and management plan identified the
following requirements for further investigation:
•Better define radon and gamma levels in the homes
•Delineate the exact perimeters of radon contamination
•Completely characterize the nature and extent of the source of the
  radon  gas
•Evaluate any potential water contamination
•Assess the level  and extent of soil contamination
•Evaluate possible uptake by garden vegetables
•Appropriately evaluate other above-background areas identified
  in the  aerial gamma survey
  In summary, the risk assessment and management plan provided
a quantification of the  potential risk associated with  exposure to
various levels of radon progeny and also provided overall direction
for both the response action and field investigation. This plan was
available as guidance to all the affected agencies and the public on
Dec.  6, 1983; the day the Superfund removal action commenced.

IMMEDIATE FIELD ACTIVITIES
Confirmation of Radon Progeny Levels

  The most immediate concern of the agencies involved  in the
response action was to better define the radon levels in the homes.
The initial  radon survey performed by NJDEP utilized activated
carbon canisters placed in the  residences for three  days. These
canisters  and subsequent  analysis  on  a sodium iodide  detector
coupled to a multi-channel analyzer measured the concentration of
radon gas  in picoCuries/Liter (pCi/1) of air.
  Since the concentration of radon progeny can vary from house to
house due to  changes in the  percentage of  radon  progeny  in
equilibrium with the radon  gas, a direct measurement of radon pro-
geny, rather than radon gas, was recommended by NJDEP and the
USEPA. To measure radon progeny directly, Radon  Progeny In-
tegrated Sampling Units (RPISU) were utilized; these units  were
developed  by Colorado State University and are the primary in-
tegrated air sampling system used by the USEPA. In operation, the
unit consists essentially of an air  pump and a sampling head located
outside of the pump housing. The sampling head contains two ther-
moluminescent dosimeters  (TLDs).
  During operation, air is  pulled through the sampling head, and
paniculate matter containing the radon decay products is trapped
on the filter. The first TLD, located in the airstream directly before
the filter,  records the alpha energy emitted  by the decay  of the
radon progeny. The second TLD records the ambient gamma
radiation which is subtracted from that measured by the first TLD.
The units were normally left in place for 5 to 7 days, at which time
the sampling heads were  removed and the TLDs  sent  to  the
USEPA's Las Vegas laboratory  for analysis on a TLD reader.
  The NJDEP sampling identified 28 homes which had abnormally
high concentrations of  radon gas.  Personnel from the USEPA's
Eastern Environmental  Radiation Facility  (EERF),  located  in
Montgomery, Alabama, conducted the initial, intensive field activi-
ty associated with the confirmation testing. USEPA-EERF person-
nel  installed RPISU's in the basements and first floor living areas
of each of the identified homes within  the first two days of the in-
itiation  of the response action; confirmation of the radon progeny
                                                                                   Table 1
                                                                    Centers for Disease Control Risk Estimates
                                                                Annual Risk of Lung Cancer for Residential Exposure
                                                                            to Selected Radon Levels

                                                             A) For Continuous LIFETIME Exposure:
                                                                  Annual Risk of Lung
                                                                               Cancer                  Radon Level
                                                                      1,000 per 100,000                  l.OWL*
                                                                       100 per 100,000                  0.1 WL
                                                                        30 per 100.000                  0.02 WL
                                                                         6 per 100,000                  0.004 WL

                                                             B) For a ONE YEAR Exposure:
                                                                  Annual Risk of Lung
                                                                               Cancer                  Radon Level
                                                                      400 per  1,000.000                  l.OWL
                                                                       40 per  1,000,000                  0.1 WL
                                                                       8 per  1,000,000                  0.02 WL
                                                                       1 per  1,000,000                  0.004 WL
                                                         • WL- Working I ocl—A wotkini loci it defined as the potential alpha entityfrom the then-lived
                                                         d«ughler» of radon which will produce I 3 • IO'MEV(MEV - Million Election Volts) ui one Iho
                                                         of air.
                                                                                    Table 2
                                                         Categorization of Residences Based on Level of Radon Contamination
                                                         Tfcr
                                                                   Radon
                                                                   Level
                                                                            Tim* Frame
                                                                                           Action (Exponra Redaction)
                                                                 over 0.5 WL   a. immediately
                                                                            (1-2 da>.)


                                                                            b within 2
                                                                            •tiki


                                                                 0.1*5 WL    Prompt (1-3
                                                                            months)


                                                                 0.024.1 WL   I-2 yean
                                                                  On: and
                                                                   below
Restrict occupancy of high lr*d artai of
home to 2 hr day Prohibn smoking in
high level areas

Temporary remedial action to get as far
below 0.5 WL as feasible using tem-
porary measures.

Temporary remedial action to gel to O.I
WL or less (Priority of action deter-
mined  by exposure level.)

Permanent  remedial action 10 reduce
exposure from non-natural source to
bclo» n CO  WL (Priority of action de-
lerrmncd by exposure level.)

No action
                                                        problem in these homes and  the  resultant classification of the
                                                        homes by Tier was attained within  the following two weeks. This
                                                        provided the data required fpr USEPA and NJDEP to authorize
                                                        and initiate the design and installation of corrective actions in the
                                                        most  severely affected residences, i.e., Tier A and Tier B  homes.

                                                        Better Delineation of the Perimeter of
                                                        Radon Contamination

                                                          The identification of a radon contamination problem in Mont-
                                                        clair and  Glen Ridge was  reported widely and extensively by all
                                                        media, leading to extreme concern  on the part of the public; con-
                                                        cern that was warranted given the limitations of the investigation at
                                                        that point. A second major objective  of the joint USEPA-NJDEP
                                                        field  investigation  was  to  identify  any additional   affected
                                                        residences. In addition, efforts were made to determine the extent
                                                        of  the radon  contamination  problem by  establishing a boundary
                                                        around the study  areas beyond which  there were no  affected
                                                        residences. To achieve this goal, a sampling program based on grab
                                                        sampling and RPISU monitoring was designed.
                                                          Grab sampling techniques measured the radon daughter product
                                                        activity captured on a  filter over a 5 min sampling period.  The
                                                        filters  were transported to  a  mobile laboratory  for immediate
                                                        analysis using alpha radiation detectors. Two grab sampling tech-
                                                        niques were utilized: the Kusnetz method and the Thomas modified
                                                        Tsivoglov method.
458
CASE HISTORIES

-------
  The CDC Health Advisory provided the following guideline for
radon concentrations in homes; "...The current USEPA limit for
radon levels in dwellings is 0.02 WL, which is at the higher end of
the range of  natural background levels found in homes in  the
United States. Houses in the United States normally range from
less than 0.002 WL to 0.04 WL, depending upon site of construc-
tion. Natural background levels above 0.01 WL are infrequent..."
This  guideline served as the basis  for formulation of the sampling
program.
  A  major premise of the study was that grab sampling  could be
utilized as a proxy to the more time-consuming RPISU sampling.
In order to be conservative,  samples were generally taken in the
basements of the residences, the  area of expected highest radon
concentration. Furthermore, as an added measure of conservatism,
a grab sample value of 0.01 WL was used to determine the need for
subsequent RPISU sampling for  5 to 7  days. A  RPISU value of
0.02  WL or greater, in turn, was used to define the need for correc-
tive  action. A schematic of the  sampling  program is shown in
Figure 1.
  The reliability of the grab sampling methodology for determina-
tion  of indoor radon progeny levels in Montclair and Glen Ridge
was assessed by staff from the USEPA's Office of Radiation Pro-
grams.' An examination of unpublished data from an 18-month
USEPA  instrument comparison  project  in Butte,  Montana,
showed that only 2.1% of 608 grab samples exhibited values of less
than 0.01 when the actual annual level was above 0.02 WL, and on-
ly 0.23% of 435 samples exhibited  levels of less than 0.01 WL when
the actual annual level was  above  0.03 WL.
  The  grab sampling  program began with the establishment  of
sampling sectors; 13 sectors were  eventually established in Mont-
clair  and 15  in Glen  Ridge.  Sectors consisted  of 10-15  homes
grouped around a central point where a mobile laboratory could be
stationed during analysis. In addition, grab samples outside of the
                           Table 3
               Radon Progeny Sampling Summary
                       (March 26, 1984)
                 >  of Homes
               Inside Sectors
                2  (Montclair)
               14  (8-Montclair)
                  (6-Glen Ridge)
               25  (15-Montclair)
                  (9-Glen Ridge)
                  (1-West Orange)
309  (83-Montclair)
    (165-Glen Ridge)
    (36-West Orange)
    (25-East Orange)
                      > of Homes
                    Outside Sectors
                                                 I of Homes
                                      (64-Montclair )
                                      (76-Glen Ridge)
                                      ( 1 7-West Orange)
                                      (4-East  Orange)
                                                      14
                                                      25
                                                     470
 Total        350  (108-Montclair)   '61 (64-Montclair)   511
 Residences        (180-Glen Ridge)     (76-Glen  Ridge)
 Monitored         (37-West Orange)     (17-West  Orange)
                  (25-East Orange)     (4-East Orange)
sectors were taken when requested by residents or by design if a
location outside of the sector would aid in establishing the desired
boundary. Grab sampling  was performed by USEPA-EERF per-
sonnel throughout December and was continued by members of the
USEPA's Field Investigation  Team (FIT), NUS Corporation, in
January and February. Eventually, over 500 homes were sampled
(Table 3). The affected areas and the perimeter of affected homes
are shown in Figures 2 and 3.
                                                          Figure 1
                                              Flow Chart—Radon Sampling Project
                                                                                                 CASE HISTORIES
                                                         459

-------
                                                                                    UNABLE TO GAIN ACCESS
                                                           Figure 2
                                           Summary of Results of Indoor Radon Sampling
                                                 (Montclair Site, Montclair, NJ)
IMMEDIATE RESPONSE ACTION

Implementation of Corrective Actions

  The design of corrective actions in Tier A and B homes was in-
itiated  immediately upon confirmation of tier status by RPISU
sampling.  The USEPA  Region II Technical Assistance Team
(TAT)  contracted the Arix  Corporation of Grand Junction, Col-
orado to design and implement the corrective action.
  In the last decade, a number of different remedial techniques
have been developed for reducing the concentration of radon  gas
and radon progeny in homes; these techniques, discussed extensive-
ly in the literature, include the sealing of cracks and crevices to
reduce radon migration into the homes, the use of passive systems
such as the  construction of subsurface ventilation  systems to in-
tercept and reroute the  radon before it enters the home and the use
of  active  systems  such  as  the  installation  of  electrostatic
precipitators to remove particulate matter from the air.  However,
                                                       given the high levels of radon in the residences in Montclair and
                                                       Glen  Ridge and the need to implement a corrective action that
                                                       would be guaranteed to be successful, a very conservative approach
                                                       to the problem was taken.
                                                          Fresh air ventilation systems were designed and installed in the
                                                       basements or crawlspaces of  all affected homes. These systems
                                                       pumped outside air into the homes at a rate averaging 200 ftVmin,
                                                       diluting the indoor air before being vented through louvers or ex-
                                                       haust fans. Individual electric heating units were included as a com-
                                                       ponent of each ventilation system and were designed to heat the
                                                       outside air to approximately 65 °F before being introduced into the
                                                       homes.  Separate electrical circuits and meters were installed to
                                                       allow direct billing of system operation and maintenance costs to
                                                       the State of New Jersey's Spill Fund.
                                                          Ventilation systems  were installed  in 22  homes during the
                                                       response  period.  These  systems  are  considered a  temporary
                                                       measure, designed to reduce the concentration of radon and radon
460
CASE HISTORIES

-------
                  TOWN MOPCHTY
                     VACANT LOT
                  UNAILE TO QAIN ACCESS
                                                           Figure 3
                                          Summary of Results of Indoor Radon Sampling
                                                (Glen Ridge Site, Glen Ridge, NJ)
progeny in the home until a permanent remedial solution can be im-
plemented. The systems have been successful in reducing the con-
centration of radon progeny in the Tier A and B  homes to Tier C
levels or lower. The concentrations of radon gas and radon progeny
in selected homes before and after installation of the ventilation
systems are shown in Table 4.
  Post-remedial monitoring continues on a quarterly basis to en-
sure that  the systems are operating properly. In  addition, Tier C
homes are monitored quarterly to ensure that the radon progeny
levels  have not  changed.  In several  instances,  the quarterly
monitoring identified Tier C residences as having Tier B levels of
radon progeny. These homes, once identified, were immediately
scheduled for  installation of ventilation systems.
  Installation  of ventilation systems in all identified  Tier A and B
homes was completed in early April, 1984, thus meeting the direc-
tives of the CDC Health Advisory.
ADDITIONAL FIELD INVESTIGATION
ACTIVITIES

Source Characterization Study

  A number of other field activities were initiated to more fully
assess the extent of the problem.
  As an initial step, the New Jersey Geological Survey conducted a
review of historic photos and maps to identify areas where con-
taminated material may have been deposited. The review identified
a number of areas where excavation or filling of material may have
occurred; several large areas were occupied by residences showing
radon contamination.
  After  the  historical  review,  a  program  was  developed  to
characterize the general nature and extent of the material believed
to be the source of the radon gas. This investigation was conducted
                                                                                                 CASE HISTORIES
                                                                                                                         461

-------
                                                            Table 4
                                    Reduction of Radon and Radon Progeny In Remediated Residences
                                            Pre-Remediation
                                                                          Post-Reined tat ion    Radon
                                                                                          »  Reduction
                                       Radon Progeny
                                      (Workinq  Level)
                         Residence A
                            Basement      0.201
                            First Floor    0.287
                            Second Floor   NA
                                               Radon
                                              (pCi/L)
                                               1 10.4
                                                80.6
                                                65.)
     Radon  Progeny
    (Working Level)
                                                                   II
0.010
0.021
NA
                                                                          12
0.012
0.027
NA
                                                                                13
               Radon
              (pCl/L)
0.006
0.005
NA
14.2   87
10.3   87
 9.4   69
                         Residence B
                            Basement      1.S49
                            First Floor    0.170
                            Second Floor   NA
                                               440.0
                                                50.0
                                               132.0
0.001
0.001
NA
0.004
0.005
NA
         0.4  99
         0.5  99
         0.5  99
                          Residence C
                            Basement      0.204
                            First Floor    0.256
                            Second Floor  NA
                                               102.0
                                                83.8
                                                92.6
0.044
0.026
NA
0.081
0.153
NA
0.017
0.0)4
0.044
36.1   65
26.a   68
25.4   73
                          Residence 0
                             Basement      0.505
                             First Floor   0.466
                             Second Floor  NA
                                               186
                                               112
                                               120
0.014
0.042
NA
0.004
O.OOS
NA
        10.0  95
        17.8  84
        14.1  88
                          NA-Not Available
by the Region II Field Investigation Team, NUS Corporation, and
consisted  of an extensive outdoor surface gamma survey using
Micro-R meters at a height of 3 ft above the ground followed by a
subsurface investigation into those areas evidencing surface gamma
anomalies. For consistency, an attempt was made to survey all
properties included  in the radon monitoring sectors.  Eventually,
209 properties in both study areas were surveyed. Significant sur-
face  gamma  anomalies  were  investigated  further  by drilling
boreholes: 326 boreholes were drilled in the neighborhood.  The
level of gamma radiation in all boreholes was determined using a
scintillometer to measure the vertical extent of contamination. Soil
samples were taken and analyzed using gamma spectrometry to ful-
ly characterize the contaminated material.
  The result of the study identified the source of the contamination
as concentrated Radium-226. It became clear that the material had
been used as fill for grading purposes. It averaged 5 ft in depth but
was generally no deeper than 10 ft. The  material was concentrated
in four large areas with scattered discrete  pockets  close-by.  The
volume of material requiring removal was estimated to be approxi-
mately 20,000 yd3. The locations of the contaminated material in
Glen Ridge and Montclair are shown in  Figures 4 and S.
Indoor Gamma Radiation Surveys

  All Tier A, B and C homes were subjected to two indoor gamma
surveys. The  first was an  exposure level survey conducted by
USEPA-EERF personnel to establish  whether residents were being
exposed to excessive levels of gamma radiation inside their homes.
  In  this  survey,  the  engineers  utilized  scintillometers  and
pressurized ionization chambers (PIC).  In  addition, a procedure
called a delta  gamma survey was performed. In this survey, one
uses  detection equipment to  measure relative  differences in the
gamma radiation  emanating through the walls and floors of the
residences; it was utilized to ascertain the location of contaminated
material under and against the foundation of the home. The results
of these surveys were represented graphically on floor plans for
each home.

Gronndwater Monitoring Program

  The CDC  Health Advisory also recommended that the potential
for  groundwater contamination be fully evaluated.  Immediately
upon the discovery of the problem, all public water supply wells in
                                                       the area were sampled for radionuclides; none showed any concen-
                                                       tration in excess of standards. There were no private wells in the
                                                       area.
                                                         Subsequent to the completion of the source characterization
                                                       study which identified the  area where  the contaminated material
                                                       was located, a program was developed to ascertain the presence of
                                                       radionuclides in the  major  groundwater systems in the area. This
                                                       program, currently ongoing, entailed the  installation  of ground-
                                                       water  monitoring wells  in unconsolidated surface deposits  to
                                                       monitor localized groundwater and in the underlying bedrock for-
                                                       mation to monitor regional groundwater flow.

                                                       SECONDARY RISK ASSESSMENTS

                                                       Evaluation of Gardening

                                                         A major public concern after discovery of the problem in Mont-
                                                       clair and Glen  Ridge was the threat posed by ingestion of garden
                                                       vegetables grown in the study area. The NJDEP's Bureau of Radia-
                                                       tion Protection conservatively estimated the potential for uptake
                                                       and ingestion of radionuclides. After consultation with CDC and
                                                       the USEPA's Office of Radiation Programs, the conclusion was
                                                       reached that, given the non-uniform and limited existence of highly
                                                       contaminated homes, the health risks from ingestion of vegetables
                                                       grown in radium-contaminated soils did  not warrant a  general
                                                       recommendation that residents discontinue gardening. However,
                                                       for those locations where the concentration of radium in  the soil
                                                       was known to exceed 100 picoCuries per gram (pCi/gm) residents
                                                       were advised to discontinue extensive gardening.
                                                         Evaluation of the source characterization study identified very
                                                       few gardens  located in any area  with  elevated  gamma radiation;
                                                       where this occurred,  the residents were so advised. NJDEP offered
                                                       assistance  in locating  gardens  and also offered to analyze any
                                                       vegetables which may have been grown  the previous year and
                                                       canned or home processed.

                                                       Evaluation of Gamma Radiation  Levels

                                                         The CDC  evaluated the surface and indoor gamma radiation
                                                       levels in Montclair and Glen Ridge. Based on this review, there was
                                                       no  need for a recommendation that access to any area be restricted
                                                       to the general public in any way. In addition, the CDC developed a
                                                       method to estimate, in a conservative manner, the annual dose to a
462
CASE HISTORIES

-------
                                                          Figure 4
                                         Soil Removal Map (Montclair Site, Montclair, NJ)
resident due  to exposure to gamma radiation. This method  as-
sumed that the residents spent the entire year confined to their pro-
perty, 18 hr indoors (8 hr in the bedroom, 5 hr in the living area, 5
hr in the basement) and 6 hr outdoors. The highest gamma reading
in each room was utilized in calculating potential exposure; out-
doors, a weighted average based on distribution of contamination
was used.
  Based on these calculations, no resident living on any of the pro-
perties in the study area would exceed the regulatory limit of 500
mrem/yr  for doses  to an individual in  the general public. Never-
theless, any residents whose calculated dose approached the limit,
even with the conservative assumptions utilized, were advised of
steps that could be taken to reduce their exposure. In addition,
residents were provided with both outdoor maps and floor plans of
their property showing the gamma levels and highlighting any area
where the gamma readings were higher than 60 /iR/hr, the gamma
radiation  level that would result in an annual exposure of 500
mrem/yr  if occupied continuously.

PUBLIC  INTERACTION
  The announcement of the discovery of radioactive material in
residential communities  resulted  in  extreme public  concern,
especially since the magnitude of the problem was not known at the
time of the announcement and could not be known without  the
planned extensive field investigation. To attempt to educate, in-
form and reassure the public,  an extensive program of public in-
teraction was developed. A combination command post and infor-
mation center was established in the Montclair Municipal Building,
and the telephone numbers were widely publicized.
  During the first several weeks of the project, state and federal
staff averaged two  meetings  every day with  township officials,
school boards, affected homeowners and the general public. All ef-
forts were coordinated with the local  health departments, which
provided extensive support in reaching  the public. In addition, the
Township of Montclair formed a task force of local citizens which
proved extremely helpful  in  disseminating information and ex-
pressing the concerns of the populace.
  The grab sampling program was explained at every opportunity,
and the citizens were informed as to the day  the sampling teams
would be in their area. This often led to residents providing keys to
their neighbors or arranging to be home during that time; this was
crucial to the success of the grab sampling effort.
  Individual access forms were obtained from homeowners prior to
the gamma survey; if drilling on that  property was required, the
homeowner was contacted in advance of the drilling date.
  When homes were identified as Tier  A or Tier B, thus requiring
corrective action, teams  of federal, state and contractor personnel
met with the homeowners to explain what was to occur and to ad-
dress any concerns.
  The request heard from the public most often was for the timely
release of data,  especially the indoor radon  monitoring results.
With this in mind, NJDEP and the USEPA attempted to contact
the homeowners by telephone whenever indoor data were available.
In the case of the grab sampling results, this often occurred on the
same day as the sampling. RPISU results for  the longer sampling
period were also transmitted by telephone upon receipt from the
laboratory. All indoor results were  confirmed by the  NJDEP  in
writing to the  homeowner.
                                                                                                 CASE HISTORIES
                                                          463

-------
                                                             Figure 5
                                          Soil Removal Site (Glen Ridge Site, Glen Ridge, NJ)
  Upon completion of the field investigation, all residents who had
work performed on their property received a file containing all per-
tinent information, specifically, indoor radon results, indoor gam-
ma radiation levels, outdoor gamma radiation levels and the loca-
tion of any  contaminated  material. Given the highly technical
nature of  the data,  the USEPA and  NJDEP  personnel  were
available on several evenings and Saturdays to meet with individual
homeowners  to review and explain their  information.

CONCLUSIONS
  Responding under the emergency provisions  of CERCLA, the
USEPA and NJDEP have  reduced the  exposure of the affected
residents to radon  gas.  In  addition, other potential health risks
have been assessed. Moreover, the  source of the contamination has
been identified. This allows the agencies to plan for and pursue a
more permanent solution, such as the removal of the  contaminated
material to an off-site  location.
REFERENCES

 1. Richardson, A.,  USEPA, Office of Radiation  Programs, personal
    communication, Feb. 24, 1984.
                                                            Bruno, R.C.,  "Sources  of  Indoor Radon in Houses:  A Review,"
                                                            JAPCA. 33. 1983,  105-108.
                                                            Colle'. R.  and McNall,  P.E.. ed., Radon in Buildings—Proc. of a
                                                            Round/able Discussion of Radon in  Buildings held at  the National
                                                            Bureau of Standards, Gaithersburg,  MD. June 15, 1979, U.S.De-
                                                            partment of Commerce, June, 1980.
                                                            USEPA, "Standards for Remedial Actions in Inactive Uranium Pro-
                                                            cessing Sites,  Final Rule", Federal Register.  48. Jan. 5.  1983, 590-
                                                            606.
                                                            Gesell, T.F., "Background  Atmospheric 222Rn Concentrations Out-
                                                            doors and Indoors: A Review", Health Physics. 45, No. 2. Aug., 1983.
                                                            Hurwitz,  H ,  Jr. The Indoor Radiological Problem in Perspective,
                                                            General Electric Technical  Information  Series,  Feb.  1981,  No.
                                                            81CRD025.
                                                            Nero, A.V., "Indoor Radiation Exposures from :~Rn and Us Daugh-
                                                            ters:  A View of the Issue", Health Physics. 45. No. 2. Aug., 1983.
                                                            Nero, V.A., "Airborne Radionuclides and Radiation In Buildings: A
                                                            Review", Health Physics, 45, No. 2, Aug., 1983.
                                                            NUS Corporation, Figures 1-5, Graphics provided courtesy of NUS.
                                                            Wadach,  J.B., Clarke, W.A. and Nitschke,  I.A.,  "Testing of Inex-
                                                            pensive Radon Miligative Techniques In  New York State Homes",
                                                            Paper presented at the Twenty-Ninth Annual Meeting of the Health
                                                            Physics Society, June, 1984.
 464
CASE HISTORIES

-------
                 LESSONS  LEARNED  IN  THE  CONDUCT OF
                           REMEDIAL ACTION ACTIVITIES

                                          M. JOHN CULLINANE, JR.
                                             RICHARD A. SHAFER
                             U.S. Army Engineer Waterways Experiment Station
                                            Environmental Laboratory
                                              Vicksburg, Mississippi
INTRODUCTION

  In response to the threat to human health and the environment
posed by uncontrolled hazardous waste sites, a variety of civilian
and military  remedial action programs have been implemented.
Prominent among these are the Installation Restoration Program
(IRP) and  the  Environmental Restoration  Defense Account
(ERDA) managed by the Department of Defense (DOD) and the
Superfund Program (CERCLA) and Resource  Conservation and
Recovery Act (RCRA) managed by the U.S. Environmental Pro-
tection Agency (USEPA).
  Cleaning up hazardous waste sites is a new field in science,
engineering and public policy. Consequently, little data are readily
available on the  performance and  costs of remedial technologies
and on management of site cleanups. Although the remediation
activities conducted by DOD, the USEPA and private sector com-
panies are similar in nature, there  is usually no formal coordina-
tion effort for transferring technology or lessons learned from one
agency to another. In addition, there is usually no formal mechan-
ism for intra-agency transfer of technology. Indeed, the technology
transfer that  does occur is  often a random happenstance process
based on word-of-mouth documentation.
  In response to this apparent communications gap, the U.S. Army
Engineer Waterways Experiment Station (WES) conducted a sur-
vey of Federal, state and private sector personnel with experience
in the conduct of remediation activities. The primary purpose of
the survey was to compile a comprehensive problem-solution data
base  that could be used  by project personnel in the conduct of
future remediation activities. Over  150 individuals with first hand
experience in the administration design and construction of reme-
dial action projects were contacted during the course of the survey.
  Although the  survey  identified  over  125  individual  lessons
learned, because  of space limitations, only the  17 most significant
are presented in this paper.  For  organizational purposes, the
lessons learned presented below are categorized into the two basic
steps associated with implementation of a remedial action: plan-
ning and design and construction.
PLANNING REMEDIAL ACTIONS
Recognizing Technological Limitations

  Problem: In some cases, the technology needed to handle the
total cleanup of a site may not exist. For example, where contam-
ination of a subsurface aquifer has occurred, it may be impossible
to flush all contaminants out of the porous geologic  units simply
because of the limited access any flushing agent has to pore space
in the units. In other instances, the reactions (adsorption, precip-
itation, etc.) used to remove a contaminant from surface  water
may not be efficient enough to restore the water to its precontam-
ination condition.
  Solution: Misconceptions about the existence of available tech-
nology and its associated capabilities are common among laymen
as well as within the engineering community. Engineers concerned
with the planning and design of remedial investigations and design
must be aware of available technologies and particularly the limi-
tation of technologies.  This awareness can  be developed through
the implementation of formal training programs.  Training can be
accomplished on an individual basis or in organized classes. Ad-
vantage should be taken of the many training opportunities avail-
able in the private sector. Further, it should  be  recognized that
technologies in this area are rapidly changing.  As a result mainte-
nance of an adequate level of competence can only  be accomplished
by the conscientious efforts of individuals.
Establishing Cleanup Criteria
  Problem: Establishing cleanup criteria, i.e.,  "how clean  is
clean" has proved to be one of the most difficult problem areas
that must be resolved. In many cases, there are no firm guidelines
or regulations limiting levels of specific pollutants in the environ-
ment. Contamination limitations are usually established on a case
by case basis in consultation with appropriate local,  state and fed-
eral regulatory agencies.
  Solution: To date, the "how clean is clean" problem has not
been resolved. The USEPA is currently addressing  the problem and
is attempting to develop risk-based guidelines for maximum allow-
able  contaminant  concentrations. Until appropriate criteria  are
established, site cleanups must continue to be addressed on a site
specific basis. The extent of cleanup  will  depend on the hazard
posed by the site as judged by four major factors:  (1) nature of the
waste, (2) dispersal pathways, (3) receptor characteristics and (4)
site management.
  In most cases, restoration of a site to a state equivalent to its pre-
contamination condition will not be practical. The relationship be-
tween cost and cleanup is an ever-steepening curve  with the final
steps to 100%  restoration the most expensive. The level of restora-
tion will usually be balanced against the cost of restoration  at the
point where immediate adverse effects to the surrounding environ-
ment are eliminated and long-term releases and dangers from bio-
accumulation  of contaminants are controlled at  some low level.
Many sites may never reach a state of restoration where the land
                                                                                          SITE REMEDIATION
                                                       465

-------
can be  designated for unlimited use. In extreme cases, on-site
contamination  may  remain at levels that require indefinite  re-
stricted access to the site.

Treatment Process Development
  Problem: In some cases, it may be necessary to collect and treat
hazardous wastes or contaminated  materials. A variety of treat-
ment processes are available for specific applications. Planners and
designers of a remedial action project may be relatively inexper-
ienced in the selection and application of hazardous waste treat-
ment processes.
  Solution: Planners and designers  must either develop some ex-
pertise in the treatment of hazardous waste or such expertise must
be obtained from outside sources. It should be remembered that
there is only limited experience with many processes being applied
to the treatment of hazardous waste. For this reason,  outside con-
sultants and contractors  are also limited in their experience with
various hazardous waste treatment processes. As a result, the con-
sultant should be questioned thoroughly and several different opin-
ions should be solicited from as many sources as practical.

Use of Industrial Specialists
  Problem: Although  the  typical remedial action project  is de-
signed as a construction project,  there are many facets of remedial
action projects  that are more appropriately handled by other than
engineering personnel. Examples include industrial hygienists,  in-
dustrial specialists,  toxicologists, fire prevention and  protection
experts, chemists, etc. Planners and designers are often reluctant to
admit a lack of knowledge and have  not effectively used sources of
specialized expertise.
  For example, after the award of  a remedial action contract at
one Superfund  site, a representative of an electric utility was
brought to the site to  discuss  the possibility of PCB contamina-
tion in electrical components other  than transformers.  After ex-
tensive discussions with the specialist, it was determined that addi-
tional pieces of electrical equipment  on the site which had not been
tested for PCB  fluids might contain  PCBs. If this information had
been known at the time of the design, the contract could have been
written  differently and the cost of removing the PCB-contam-
inated components from inside the other electrical items could have
been included in the basic contract  rather than in a change order
as ultimately required.
  Solution: Planners, designers and  constructors should be trained
to recognize the need for specialized areas of expertise during all
phases of  the  remedial  action project.  The input of specialists
should be incorporated into  the design process as early as pos-
sible.

Limited Knowledge of Regulations
   Problem: Engineers responsible  for  planning  remedial  action
projects are often relatively unfamiliar with laws, regulations and
procedures. Furthermore,  local, state and federal  requirements
may conflict. For example, federal regulations define  transformers
as uncontaminated if PCBs are less than 50 ppm. Thus, under fed-
eral requirements these  transformers can be landfilled at any
approved non-secure landfill. However, some states (Pennsylvania,
for example) do not allow any transformers to be disposed of in the
State. Thus, on one Superfund project noncontaminated material
had to be transported  out of state to a secure landfill at substan-
tially increased  cost.
  Solution: Planners and designers must be thoroughly familiar
with all applicable laws, regulations  and procedures. This familiar-
ity must extend to local, state and  federal requirements. Selected
remedial action alternatives must meet the most stringent require-
ments. Conflicts in regulations and  laws  must be resolved early in
the planning process. Where uncertainty exists, regulatory agencies
should be consulted for  interpretation of requirements. It should
be noted, however, that USEPA  interpretation of regulations tends
to be non-uniform.
                                                        Permits, Licenses and Approvals
                                                          Problem: Failure to recognize the major  time constraints that
                                                        may be involved in obtaining necessary permits, licenses and ap-
                                                        provals can cause extensive time delays.
                                                          Solution: It must be recognized that  the time from project for-
                                                        mulation  to project implementation may be significantly increased
                                                        by the necessity to obtain appropriate permits and approvals from
                                                        local, state and Federal regulatory agencies. The exact nature of re-
                                                        quired permits,  licenses and approvals will depend  on  the charac-
                                                        teristics of the site, location of the site and type of remedial action
                                                        selected for implementation.  All permits, licenses  and approvals
                                                        should be obtained prior to start of the remedial action. Although
                                                        little can  be effectively done to speed up the regulatory process, a
                                                        detailed knowledge of required approvals will prevent unnecessary
                                                        delays once on-site work is initiated.

                                                        Community and Public Relations
                                                          Problem: The importance of community relations during all re-
                                                        medial activities cannot be overstressed. Good community relations
                                                        can ensure that a  project will  proceed smoothly  and efficiently
                                                        from start to finish. Poor community relations can result in project
                                                        delays, increased costs and residual resentment in the community.
                                                        Engineering personnel tend to think of hazardous waste sites as a
                                                        purely technical problem and do not understand the importance of
                                                        good community relations.
                                                          Solution: Abandoned hazardous waste sites and spills of hazard-
                                                        ous materials are not simply a technological problem, but also have
                                                        political,  economic, psychological, social and human health im-
                                                        pacts as well. There are good reasons why people are likely to be
                                                        highly concerned about hazardous waste problems and  proposed
                                                        cleanup efforts. Unless community relations are arranged with care
                                                        and skill, there can be a tense, agitated public looking  for help,
                                                        but unsure where to turn and likely to be suspicious of any re-
                                                        sponse that seems to be half-hearted.
                                                          Recognizing that hazardous waste sites are more than just a tech-
                                                        nical concern,  the USEPA has developed  community relations
                                                        guidance  based on the experiences of USEPA regional offices in
                                                        handling  both hazardous waste  remedial action projects and haz-
                                                        ardous materials spills. Interim community relations guidance has
                                                        been issued requiring development of a community relations plan
                                                        for all hazardous waste sites where Federal funds will be spent for
                                                        more than two weeks.
                                                          The plans require that a substantial level of effort be devoted to
                                                        interacting with local communities at each site. The amount of in-
                                                        teraction  is determined by projecting both  the degree of citizen
                                                        concern and the environmental problem at the site. The more vis-
                                                        ible and serious the hazardous waste site, the more active the com-
                                                        munity relations program  should be.
                                                          The USEPA attempts to handle the concerns and expectations of
                                                        local communities  with foresight, care  and compassion. The pri-
                                                        mary purpose of implementing a community relations policy is to
                                                        assure  that  actions  at hazardous  waste sites are understood,
                                                        accepted  and supported by local communities that may be affected
                                                        by the site. The policy stresses the importance of carrying out
                                                        cleanup actions without disrupting the normal life of the commun-
                                                        ity.
                                                          A well  thought out program of community relations is an integral
                                                        element of any strategy to achieve cost effective solutions at re-
                                                        medial  action sites. The problems at some hazardous waste sites
                                                        may be difficult to understand and occasionally frightening to the
                                                        uninformed. Unless the concerns of communities affected by the
                                                        site are understood and addressed, it is possible that resulting mis-
                                                        understandings will cause  long delays and cost overruns.


                                                        Statements to the Media

                                                          Problem: Remedial action investigations and projects attract the
                                                        attention of the media. From time to time,  both government and
                                                        contractor personnel will be requested to make statements to the
 466
SITE REMEDIATION

-------
media. Comments delivered in jest to relieve the tension of a par-
ticularly hazardous operation or off-hand comments not based on
fact may lead to unnecessary concern on the part of the media and
a sensational story. For example, when asked how he would handle
shock sensitive material, one contractor replied, "...we will treat it
just like nerve gas...." A poor analogy because the only words the
press ever heard were nerve gas.
  Solution: Statements to the media should follow the simple rule:
"Be careful what you  say and say what you  mean."  If at all pos-
sible, avoid dealing directly with the press. A Public Affairs Officer
(PAO) should be on-site whenever reporters or media are present or
expected. If the press contacts you, first refer them to the PAO for
information. For support on technical questions, the PAO should
have a technical representative participate in press interviews.

DESIGN AND CONSTRUCTION OF REMEDIAL ACTIONS
Contractor Responsiveness

  Problem: The conduct of remedial action projects requires spe-
cial equipment and, more importantly,  special skills by contrac-
tor personnel.  It is obvious that it is in the best interest of all con-
cerned to utilize only competent contractors.  The problem is in de-
termining how contractor competence will be judged.
  Solution: The contractor's competence is usually  judged from
the bidding documents received. In addition  to traditional bid
documents (bid  form, bid bond, performance bond, etc.), the
bidding documents on several Superfund projects have required
the bidder to  submit  the  following additional information with
his proposal: the contractor's prior experience in removal of haz-
ardous  wastes, the qualifications of the contractor's personnel
who will perform the work, the contractor's equipment that will
be available for performing the work, agreements with transporters
to haul  the hazardous waste and agreements with  disposal sites to
handle the various hazardous wastes. These documents are used to
determined the responsiveness of the bidder.
  The use of  these documents is illustrated in the following ex-
ample from a  Superfund project. The disposal site for waste ma-
terial containing greater than 500 ppm  of PCB proposed by the
two lowest bidders was only licensed to store the material. Con-
tract documents required that all waste material be shipped  direct-
ly to  the site of final disposal. The two lowest bidders were de-
clared to be non-responsive. The lowest bidder challenged this de-
cision. The court upheld the original decision. A better scheme for
ensuring the competence of contractors is to prequalify all bidders.
Contractor Pre-Bid Site Access

  Problem:  A contractors pre-bid access to  the site will probably
be  limited to  any pre-bid  conferences.  Such limited site  access
makes it difficult for the contractor to develop a reasonable  under-
standing of project complexities and may result in high bids based
on the contractors uncertainties.
  Solution: Aerial or  typical ground level photographs of the site
should be included in  the bid documents. Care should be taken to
ensure that  selected photographs  are representative of site con-
ditions. Appropriate disclaimers should be  placed on the photo-
graphs so that the final interpretations are the contractors'.
Utility Relocation

  Problem:  Potential remedial action sites  may  contain  various
utility services such as telephone, power, water and sewer lines.
These utilities may serve the site or merely pass through the site on
utility easements. Utility companies may be  slow to relocate serv-
ices resulting in potential project delays and added project costs for
the relocation. The response of utility companies  may be particu-
larly slow at remedial action sites.
  Solution: The site should be inspected by project designers early
in the remedial investigation process. Particular attention  should
be given to  identifying  utilities that may require relocation. The
identification of underground utilities is especially important. This
determination  of utility corridors can be accomplished by either on-
site inspections or review of old maps, plans, etc. Once utility serv-
ices  have been identified,  coordination with utility companies
should be initiated as soon as practicable. Required utility reloca-
tions can be accomplished by the remedial action contractor or by
the utility. The nature of the site (degree of hazard) will probably
determine the manner in which the relocation is  accomplished.
The decision should be made in consultation with the appropriate
utility companies.

Project Phasing and Scheduling

  Problem:  Climate and weather conditions may severely impact
the conduct of remedial action  activities. Projects  have been de-
layed by the designer's failure to consider these factors. This prob-
lem may be particularly acute in those areas of the country where it
is  customary to shut down construction projects  during winter
months.  For example, at one remedial action project  the waste
treatment plant had just been started when the contractor assumed
that it was time to shut down  for the winter;  project designers
had  assumed that the contractor would work through the winter.
In addition, summer heat will slow down all operations that must
be performed in protective clothing.
   Solution:  Since remedial action projects are generally designed
as a series of construction activities, the potential impacts of cli-
mate and weather related  factors must be incorporated  into the
project planning process. Project designers must be aware of local
(site specific) weather and climatic  characteristics and  develop
project schedules accordingly. The phasing of work and  how the
time of year will affect work schedules must be considered. In addi-
tion to climate and weather, consideration must also be  given to
such cyclic factors as inundation by flood waters.

Split Contracts
  Problem:  Large projects may be divided into phases with differ-
ent contractors responsible  for various phases of the project. A
typical example from  the Superfund Program is the use of one
contractor for  initial site cleanup, inventory of waste materials
and repackaging of any hazardous materials found to be contained
in deteriorated drums or packages. A second contractor may  be
used to load and haul the materials to the final disposal site. As a
result,  there may be extensive duplication of effort in  testing to
determine what materials are being handled. The split responsibili-
ties  between contracts resulted in time delays and a split in the
responsibility for ultimate disposal of the waste.
   Solution:  Where possible, all construction  activities associated
with the remedial  action should be  accomplished under  a single
contract. The prime contractor should have sole responsibility for
completion of the remedial action from initial site organization to
ultimate disposal of the waste materials. Subcontractors may  be
used by the  prime contractor; however,  the government should
have a single point of contact with  the prime contractor. If it is
necessary to have more than two contractors on-site at the same
time, they should work for the same agency.

Inadequate Design Development

   Problem: Several  remedial action alternatives  are quasi-experi-
mental in nature, i.e., they have not been fully field tested. In many
cases,  the selected alternative attempts to go directly from the lab-
oratory to the field. Problems of this nature seem to be particu-
larly acute if the solidification/stabilization alternative is  selected.
Although the solidification/stabilization alternative has been used
extensively,  the  actual  chemistry of  solidification/stabilization
technology remains more art than science. Several problems have
been reported with field scale solidification/stabilization, primarily
concerns over  the adequacy of reagent mixing  and the  potential
fire and/or explosion problem  if reactive solidification/stabiliza-
tion reagents are added with incompatible wastes. Two minor ex-
plosions and fire incidents with resultant evacuations have been re-
ported on remedial  action projects. Both  of these incidents were
related to the use of quicklime  (calcium oxide)  and the heat gen-
                                                                                                SITE REMEDIATION
                                                           467

-------
crated during the hydration process when quicklime is added to a
liquid or sludge.
  Solution: Remedial action alternatives incorporating the  addi-
tion of  chemical  reagents should be fully evaluated in the labor-
atory and field  environment. Particular attention must be given to
potential heat of reaction problems when reactive chemicals are to
be added to the waste. The potential for explosion, fire  and re-
lease of volatile organics should be evaluated. The scenario selected
for field scale addition of the solidification/stabilization reagents
should be subjected to  full scale testing with actual construction
equipment.  Problems should be identified and  corrected prior to
initiating the contract for site cleanup.

Extent of Hazard Revision
  Problem: As a project progresses, the extent of hazard  asso-
ciated with a specific site is subject to change. The degree of actual
hazard may increase or decrease.
  Solution: A preliminary judgment of the extent of hazard is gen-
erally made on any hazardous waste site selected for remedial ac-
tion. As additional data become available, the  hazard assessment
should be updated based on new field and laboratory data.  Con-
tract provisions should allow  for changes in the level of hazard.
Revised hazard estimates may be used  to adjust safety planning
and to refine designs for treatment and/or containment.
Health versus Safety
  Problem: As the required level of personal protection increases,
health and  safety requirements may conflict.  Heavy equipment
operation may  be particularly hazardous at high levels of personal
protection. It has been suggested that real heavy metal contamina-
tion is when a D-8 bulldozer or 20-ton dump truck runs over you.
  Solution: Remedial action activities should be conducted at the
lowest possible level of protection consistent with both  health and
safety concerns. Sound judgment must  be utilized  to evaluate the
trade-offs between increased  personal  health  protection  and in-
creased safety  hazards created when equipment operators are re-
quired to wear  such equipment. Two examples of this concept are:
at low  contaminant levels, full face masks for heavy  equipment
operators may not be justified since they hinder their vision; respir-
atory protection may not be necessary for truck drivers if they are
required to remain in their trucks.
Hazardous Waste Manifest
  Problem: Remedial action alternatives may include excavation
and/or removal and off-site disposal of hazardous waste materials.
                                                         Off-site disposal options require the preparation of a Hazardous
                                                         Waste Manifest. Under RCRA, an individual, firm or agency must
                                                         be identified as the generator, thus accepting responsibility under
                                                         the requirements of RCRA for the waste even after it is disposed of
                                                         in a secure landfill. Several Superfund projects have specified that
                                                         the remedial action contractor sign the manifests as the generator.
                                                         Understandably, contractors  are reluctant to  sign these "Haz-
                                                         ardous Waste Manifests" as the generator because of the inherent
                                                         liabilities. Project delays may  result because of confusion over this
                                                         point.
                                                           Solution: The person, firm or agency that will sign the "Haz-
                                                         ardous Waste Manifest" must be clearly defined in  the bid docu-
                                                         ments.  The designated generating  agency must have an on-site
                                                         representative during the accomplishment of the remedial action.
                                                         If the "performing agency" is different from the designated "gen-
                                                         erating agency," the  "generating agency" may wish to  designate
                                                         the "performing agency" as its agent for purposes of signing re-
                                                         quired manifests.

                                                         CONCLUSIONS
                                                           When reviewed in retrospect, many of the lessons learned as a
                                                         result of this study appear  to be "common sense" items.  How-
                                                         ever, each  of the problems identified during this survey actually
                                                         occurred, with many resulting  in significant project delays and cost
                                                         increases. The lessons learned data base for the conduct of reme-
                                                         diation activities is increasing as the number of completed remedial
                                                         action projects increases. The development of the level of expertise
                                                         required for efficient conduct  of remediation activities can only be
                                                         accomplished if the lessons learned by individuals are documented
                                                         in a manner providing ready access to the relevant data without an
                                                         undue expenditure of resources.
                                                         REFERENCES
                                                         1. Office of the Chief of Engineers, Preliminary Guidelines for Selection
                                                           and Design of Remedial Systems for Uncontrolled Hazardous Waste
                                                           Sites, Draft Engineer Manual EM 1110-x-xxxx. Washington, DC, 1983.
                                                        2. USEPA,  Handbook  for  Remedial Action at Waste  Disposal Sites,
                                                           EPA-625/6-82-006,  Municipal Environmental Research Laboratory,
                                                           Cincinnati, OH, 1982.
                                                        3. USEPA, Handbook for Evaluating Remedial Action Technology Plans,
                                                           EPA-600/2-83-076.  Municipal Environmental Research Laboratory,
                                                           Cincinnati, OH, 1983.
                                                        4. USEPA,  Community Relations in Superfund: A Handbook, HW-6
                                                           Interim Version, Office of Emergency and Remedial Response, Wash-
                                                           ington, DC, 1983.
 468
SITE REMEDIATION

-------
      FEASIBILITY STUDY FOR  BURIED ARSENIC WASTE
                           AT PERHAM, MINNESOTA

                                    CATHERINE E. JARVIS
                                   WILLIAM E. THOMPSON
                                   PEDCo Environmental, Inc.
                                   Waste Management Division
                                         Cincinnati, Ohio
                                          JERRY RICK
                        Twin City Testing and Engineering Laboratory, Inc.
                                        St. Paul, Minnesota
                                        DEB McGOVERN
                               Minnesota Pollution Control Agency
                                     Minneapolis, Minnesota
BACKGROUND

 In the 1930s and 1940s, the U.S. Department of Agriculture used
arsenic as a pesticide to control grasshoppers. The State of Min-
nesota provided mixing stations, storage areas and distribution
points for the grasshopper bait, one of which was in Perham. When
the program was discontinued, leftover arsenic (most likely in the
form of crude arsenic, arsenic trioxide, and sodium, calcium or
lead arsenate) was buried at a depth of about 7 ft at the Perham
fairgrounds.
                      H.I  „„ i«t,
  The Perham arsenic burial site is located in Otter Tail County,
Minnesota, at the southern edge of the town of Perham (Figs. 1 and
2). The site, which lies between a county fairgrounds cattle shed
and the Hammer's Construction building, is a fenced area with
dimensions of approximately 40 ft x  100 ft. A state highway is
located about 30 ft west of the site.
  Soil and groundwater in the immediate vicinity of the buried
waste are contaminated with arsenic. Arsenic concentrations in soil
samples taken from five locations in the trench range from 150 to
                                           .
                               '- \.'7S. I
                               V 'k.
                                              Figure 1
                                         Regional Location Map
                                                                          SITE REMEDIATION
                                             469

-------
1
s
i
1
X

_
in nil \
n \
U 1! *
1 1 1
1 ll *" ^
LOCK c»nit "^^----_^- 	 	
~_ " i wnii | \ »nu] 1 ]
: \JBUILOING X tUllOfNC i i
• • gi ' a! ITCT-lA J I .-
*"- HAHMtH'S «LL ^
HAMHfS'S CONSTDUCTION-HtTAl Bl RDIHG 1
BLACKTOP DRIVE
10D t 	 	 , 	 J_
1 Htm 1 f J
luujifK «
ILMOOP MIKING
t
I

I
Q>o«l STHT10II
f /^' | *ZO
ui A^
Tt>'^-^'/
>'"'/ //r
^'S >*
'"" \ *
^— -'---'-"'
Ltotm>
TCT-j ^ MSŁ«IC IU«I»L SITt
A WNITOIIKC KILL
CT-4 1
1
""""j
0 100ft
SCU.E
                           Figure 2
          Site Vicinity and Monitoring Well Location Map
                  i mir^T 11 p 'ijTll1: JTTiTT: ff:H'niLTITTnT3T
                                                                 12,600 mg/kg. A zone of contaminated soil with arsenic leveb ex-
                                                                 ceeding 70 mg/kg extends downward to about 14 ft, and concentra-
                                                                 tions up to 30 mg/kg have been measured in soil core samples taken
                                                                 at a depth of 19 ft. Background levels of arsenic in soil around the
                                                                 fairgrounds are less than 1.0 mg/kg. Soil with arsenic concentra-
                                                                 tions at or above 500 mg/kg is considered hazardous according to
                                                                 Minnesota regulations.
                                                                    Groundwater  contamination from  the  buried  waste has been
                                                                 monitored since 1972,  when 11 people were poisoned from a well
                                                                 constructed  near the edge of the buried waste (Hammer's well);
                                                                 arsenic concentrations up to 21 mg/1 were measured in water from
                                                                 that well.  In 1980, water from wells down-gradient from the buried
                                                                 waste (about 350 ft from  the trench) showed arsenic concentration!
                                                                 up to 0.12 mg/1. Data gathered in 1980 indicated groundwater con-
                                                                 centrations which exceeded the drinking water standard (Fig. 3).
                                                                 Background concentrations in area groundwater are less than 0.001
                                                                 mg/1.
                                                                    The burial site  was covered with a plastic liner and a day cap in
                                                                 1982. More recent samples (taken in  1983 and 1984) show lesser
                                                                 concentrations  in several of the wells down-gradient from the
                                                                 trench.  In fact, most of the recent samples are less than the drink-
                                                                 ing water  standard. The  chronological changes in arsenic concen-
                                                                 trations in the wells surrounding the burial site are shown in Figure
                                                                 3.
                           Figure 3
  Chronological Changes of Arsenic in Groundwater from 1980 to 1984
                                                                                                               1 - V» if/kg USOIIC
                                                                                                               soo >9/t« usaic
                                                                                   Figure 4
                                                               Estimated Boundaries of Contaminated Soil Excavation
                                                                  (Arsenic >• 500 rag/kg). Assuming Vertical Walls.

                                                          The Minnesota Pollution Control Agency (MPCA) hired Twin
                                                        City Testing (prune contractor) and PEDCo Environmental, Inc.
                                                        (PEI, subcontractor) to conduct a remedial investigation/feasibility
                                                        study for the Pcrham site.  Twin  City Testing conducted the
                                                        remedial investigation and issued a report (June 30, 1984) describ-
                                                        ing  the site; PEI performed  the  feasibility study. In a separate
                                                        report, PEI made a qualitative assessment of technologies for
                                                        remedial action at the Perham site. Each technology was described
                                                        with respect to environmental effectiveness, feasibility/applicabil-
                                                        ity and relative costs.
                                                          As a result of that evaluation, four alternatives were chosen and
                                                        approved by MPCA for more detailed evaluation:
                                                        1 .   Increased groundwater monitoring
                                                        2.   Groundwater pumping and treatment
                                                        3a.  Excavation and landfilling of all arsenic-contaminated soil
                                                        3b.  Excavation and landfilling  of hazardous soil only (i.e., that
                                                            soil with arsenic levels 2 500 mg/kg)
                                                        4a.  Excavation and landfarming of all contaminated soil
                                                        4b.  Excavation and landfarming  of nonhazardous soil  only and
                                                            landfilling of hazardous soil
                                                          PEI  evaluated  these alternatives  in detail,  recommended one
                                                        alternative  and developed a  conceptual design for  the  recom-
                                                        mended alternative.
470
SITE REMEDIATION

-------
                Table 1
Evaluation of Remedial Alternatives at Perham


Evaluation factors
1. Special engineering
considerations


2. Environment^ impacts
Air

Surface water


Groundwater

Soil


Overall impact on
community



3. Operation, mainte-
nance, and monitoring

4. Time required for
implementation
5. Offsite disposal
needs and transpor-
tation plans






6. Legal constraints






7. Safety requirements




8. Ease with which it can
be phased into indi-
vidual operable units

9. Implementability and
reliability





10. Total coits
Alternative 1
Increased
groundwater monitoring
None. Install 5 additional
monitoring wells



None

None


Existing contamination would
remain and could increase
Some contamination of
soil would remain and
could increase
Low. Impact is minimal as
long as contaminated soil and
groundwater remain in place
and further migration does
not occur
Quarterly monitoring of 10
wells (sampling and analysis
for arsenic)
Short, less than 2 weeks

None








None






None, other than normal pre-
cautions taken during well
installation and when
sampling water containing
arsenic
Easily divides into quarterly
sampling, analysis, and re-
porting

Immediately and easily imple-
mented. Moderately reliable.
It is unlikely that a slug of
highly contaminated gruundwa-
ter would pass beyond the 10
wells without being detected.

15,000
Alternative 2

Groundwater pumping and treatment
Considerable bench tests of treatment technologies
would be necessary before design and implementation
could be done. Pumping tests would also be
necessary before designing the well system.

Slight. Some ventilation of arsine is possible; how-
ever, amounts would be very small.
Slight to moderate. Treated or untreated groundwater
containing arsenic at or below the drinking water
standard would be released to surface waters
Minimal. This alternative reduces the arsenic in
groundwater below the site
Some contamination of soil would remain and could
increase.

Low. Arsenic wastes remain buried. Collected ground-
water is discharged at or below the standard.



Considerable attention (labor) must be devoted to
proper operation and maintenance of a pumping and
treatment facility.
Short to moderate for a package treatment plant (less
than 2 months)
Considerable volumes of sludge from effectively sof-
tening the groundwater would be be generated. At It
solids, up to 9600 gal of sludge would be generated;
at 12% solids, about 3200 gallons would be generated.
The sludge will contain arsenic, but it is not known
whether it would be hazardous. If it is hazardous,
disposal at Peoria Land Disposal is recommended. If
it is not hazardous, it might be land-applied locally
with sewage sludge.
Surface discharge would require a permit and compli-
ance with the Clean Water Act (40 CFR 401). If
sludges are hazardous, the treatment facility would
become a generator. Disposal would have to be at a
RCRA-permitted site. It must not remain on site for
more than 90 days to avoid the need for obtaining a
RCRA permit as a treatment, storage, disposal facility.
Moderate. Only normal safety procedures for handling
the treatment chemicals



Can be somewhat divided into units. Pumped ground-
water could be stored before being treated and before
being discharged to surface water or groundwater
after treatment.
Some bench testing would be required before full-scale
design and implementation. Reliability of precipita-
tion/coagulation processes in removing low-level
arsenic concentrations is not well demonstrated.
Failure to achieve necessary arsenic removal could
result in violation of permit conditions and wasted
expenditures
361,000
                                                   SITE REMEDIATION
471

-------
                                                                Table 2
                                               Evaluation of Remedial Alternatives at Perham
tvaluat ion f,u t»r\>
1.

2.













3.



4.


5.


6.

7.
8.


9.


10.
Special engineering
consideration!
Environmental impacts
Air



Surface water

Groundwater


Soil
Overall impact to
community

Operation, mainte-
nance, and monitoring


Time required for
implementation

Offsite disposal
needs and transpor-
tation plans
Legal constraints

Safety requirements
Ease with which it can
be phased into indi-
vidual operable units
Implementability and
reliabi lity

Total costs
Alternative 3a
fncdv.it inn and landfllling of all contaminated soil
None. Excavation of IbOO yd1 about 20 ft deep Is routine.


Possible release of arsenic as fugitive dust during excavation and handling, also
small amounts of arsine gas. Oust suppression techniques such «s light sprays of
water on the working face would minimize paniculate emissions. Personal protective
equipment would protect workers from arsine exposure.
Possible contamination of surface water from runoff from rainfall during the excava-
tion process. Can be prevented by excavating during dry weather.
Existing groundwater contamination would remain; however, the source of possible ad-
ditional arsenic from the waste would be removed, resulting in eventual (long-term)
reduction ot arsenic in the groundwater.
Contaminated soil would be removed and replaced with clean fill.
Possible negative impacts are air and surface water contamination. Positive impacts
include removal of a hazardous waste and preclusion of additional groundwater contam-
ination.
Operation is straightforward and simple. The only maintenance required after the
backfill is in place is revegetation of the excavated area. Air monitoring should be
conducted during excavation. Continued groundwater monitoring Is advisable for some
time after excavation.
Short. Excavation can be completed within 2 weeks. Transport and disposal would take
another week. Peoria Land Disposal requires 1 month lead time to approve waste for
their landfill.
The closest approved landfill is about 300 miles away; others are 400 or more miles
away. Transport of the 1800 yd1 of soil would be by the selected disposal facility
or by a local approved transporter.
Waste would have to be manifested and sent to an approved landfill to comply with
RCKA.
Personal protective equipment may be necessary for those Involved in the excavation.
Operation can be divided into excavation, packaging in drums or in bulk trucks, trans-
port, and disposal. Excavation and packaging must be done together. Transport and
disposal can be done as units separate from excavation and packaging.
The excavation and disposal are easily implemented and are very reliable. Failures
could include leaving some arsenic-contaminated soil in the ground or spilling mate-
rial as it is packaged or transported, all of which can be minimized or eliminated.
398.000
                                                               Table 3
                                              Evaluation of Remedial Alternatives at Perham
         Evaluation factors
                                                                                  Alternative 3b

                                                                    Excavation  and disposal of hazardous soil only
          1.   Special engineering cuns iUeratiuns



          2.   Environmental impacts

              Air




              Surface water


              Groundwater
                                           Minimal.   Excavation of 200 yd  about 6 feet deep is routine,  will
                                           require iiuny  soil core samples and analysis to define perimeter of area
                                           requiring  excavation (arsenic >SOO ppm).
                                           Possible  release  of  arsenic  in fugitive dust and small amounts of arsine
                                           gus  during  excavation.  Dust control Measures such as light sprays of
                                           water  on  working  face will minimize dust.  Personal protective equipment
                                           will reduce possible exposure to arsine.
                                           Possible  contamination  from  runoff during excavation.
                                           eliminated  by  excavating during dry weather.
This can be
                                           Existing  groundwater  contamination would remain.  Because the most
                                           concentrated  source of arsenic  (the highly contaminated soil) would be
                                           removed,  little  further  leaching, into groundwater should be minimal.
                                           Replacement of the plastic  liner and clay cap will reduce the likelihood
                                           of further leaching from less contaminated soil  remaining in place.
472
SITE REMEDIATION

-------
                                                 Table 3 (continued)
                                    Evaluation of Remedial Alternatives at Perham

Evaluation factors
Soil
Overall impact to community
3. Operation, maintenance, and
monitoring
4. Time required for implementation
5. Offsite disposal needs and trans-
portation plans
6. Legal constraints
7. Safety requirements
8. Ease with which it can be phased into
individual units
9. Implementability and reliability
10. Total costs
Alternative 3b
Excavation and disposal of hazardous soil only
Highly contaminated soil would be removed and disposed of at an approved
landfill; lesser contaminated soil would remain in place.
Minimal negative impact from possible release of arsenic into the air
during excavation. Moderate positive impact due to removal of hazardous
waste from area.
Minimal. Excavated area will have to be backfilled, regraded, and
recapped. Continued groundwater monitoring is advisable for some time.
Short. Excavation can be completed in 1 week. Transport and disposal
would require up to a few days. Peoria Land Disposal requires 1 month
lead time to approve waste for their landfill. Initial soil core monitor-
ing to define the area to be excavated would also require lead time of
about 2 weeks.
About 200 cubic yards of material would require disposal at an approved
landfill. Transportation would be by the selected disposal facility or
by a local approved transporter. The nearest disposal facility is about
300 miles from Perham.
Hazardous waste would have to be manifested and sent to an approved
landfill to comply with RCRA. The nonhazardous soil remaining at Perham
is not regulated.
Per&uiuil protective equipment may be necessary for those involved in the
excavation.
Operation can be divided into soil core sampling and analysis,
excavation, transport and disposal, and site closure (backfilling,
grading, and recapping). Site closure should immediately follow
excavation.
The excavation and disposal are easily implemented and are very reliable.
Failures could include leaving some hazardous soil inplace that could
contribute to further groundwater contamination; spilling hazardous soil
as it packaged or transported; and generation of contaminated runoff
during excavation. Measures to minimize failures include thorough
characterization of the hazardous area's dimensions from soil core
sampling and analysis; careful handling practices; and conducting
excavation during dry weather. uu«.nnM
$93,000
                                                      Table 4
                                    Evaluation of Remedial Alternatives at Perham
    Evaluation factors
                 Excavation and landfarming of all contaminated soil
1. Special  engineering
    considerations

2. Environmental  impacts

     Air


     Surface water



     Groundwater
     Soil
    Overall impact to
     community
 Considernble engineeing  is required to complete demonstration plan and program for a
 RCRA permit and for facility design and operation.
Possible  release of arsenic as fugitive dust and of small amounts of arsine during
excavation and landfarming.  Dust emissions can be minimized by water sprays.

Possible  contamination from runon and runoff from contaminated soil could be
minimized by excavating during dry weather and by controls at landfarm sit*, such
as site slope, berms, and runoff collection basins.

Existing contamination of groundwater would remain, but it may decrease over the  long
term because the source 'if arsenic would be removed.  Possibility for groundwater
contamination at the landfarming site is minimal due to immobilization in soil  and
dilution of waste over a larger area.

Existing contaminated soil  would be removed and spread over a larger area.   This
would contaminate a larger amount of soil, but the concentration would be lower.
Contamination of additional  soil  would be essentially permanent.

Negative impact includes possible air emissions during excavation and landfarming  and
contamination of the landfarming plot with arsenic.  Positive impacts include  removal
of a hazardous waste at Perham and conversion of the highly concentrated arsenic  into
a nonhazardous concentration through landfarming.
                                                                                            SITE REMEDIATION
                                                                                                                          473

-------
                                                         Table 4 (continued)
                                             Evaluation of Remedial Alternatives at Perbam
            Evaluation  factors
                                                  Excavation and  landfarming of all  contaminated soil
        3. Operation, mainte-
            nance,  and monitoring
        *,. Time required for
            implementation
        5. Offsite  disposal
            needs and  transpor-
            tation  plans

        6. Legal constraints
        7. Safety  requirements
        8. Ease with which It can
            be phased into indi-
            vidual operable units

        9. Implementability and
            reliability
                        Considerable.  Operation of demonstration program and full-scale landfarming require
                        engineering design, land preparation, water control  measures, waste applications,  and
                        vegetation of plot.  Maintenance  of vegetative cover and water control measures  are
                        required.  Monitoring needs include sampling and analysis of air, soil, soil pore-
                        liquid,  runoff, and groundwater.

                        Considerable, about a year or more.  The demonstration plan and program take from
                        several  months to a year.   Full-scale landfarming would take less than a month,  but
                        monitoring for a year afterwards  would probably be required.

                        Transport should be by a '.ocal approved hazardous waste transporter.  No offslte
                        disposal  is anticipated unless large amounts of contaminated runoff are accumulated
                        that  cannot be reapplied to the site.  This would have to be treated.

                        The landfarming site would have to be permitted under  RCRA.  This requires  substan-
                        tial  planning, engineering, and monitoring and will  take some time to  obtain, but
                        it can be done.

                        Personal protective equipment (Level C) Is recoomended.  Air monitoring should alert
                        working  personnel  of the need for additional or for  less protection.

                        Can be divided into excavation of a small  amount  of  soil for the demonstration progran.
                        followed by complete excavation, transport,  landspreading, revegetation. monitoring,
                        and closure.

                        If the demonstration program is successful  in proving that the  arsenic  is immobilized,
                        then  the application for a  permit should be  approved and the full-scale operation
                        should be reliable.
10.
Total
costs
(128.000
                SAMPLES  FROM TRENCH
                                    o
                                              •TRENCH  LIMITS
   10-
                                             MtSENIC CONCENTRATION

                                                (OOmg/IIKt

                                                70-800 «0/Mt«r

                                                l-70 mj/llc.f
                           Figure 5
    Vertical Soil Profile C-C1 with Arsenic Concentration Contours
RESULTS
  The evaluation of each of these alternatives is shown in Tables 1
to 5 while the costs are summarized in Table 6. These costs are ±
20%; no contingencies have been included.
  Option 1, increased groundwater monitoring, would cost about
$15,000 for the first year, assuming installation of four additional
monitoring wells, and about $4000 per year thereafter for quarterly
monitoring. Option 2, groundwater pumping and treatment, is the
most expensive alternative costing in excess of $360,000. Option 3a,
                                                          excavation and landfilling of all the contaminated soil, would cost
                                                          about $394,000 based on disposal at a landfill in Peoria, Illinois.
                                                          Other landfills charged considerably more, so this price represents a
                                                          minimum;  the cost could range up to $500,000 depending on the
                                                          site selected. Option 3b, excavation, landfilling  and recapping at
                                                          Perham, would cost about  $90,000.  This is the second most
                                                          economical alternative. Option 4a, excavation and landfarming of
                                                          all the contaminated soil, would cost about $128,000. Option 4b,
                                                          excavation of all the soil, disposal of the hazardous soil and land-
                                                          farming of the nonhazardous soil,  costs about the same, approx-
                                                          imately $124,000.
                                                            The information summarized in Tables 1  through 5 is rated in a
                                                          semiquantitative manner in Table 7 by assigning numbers to each
                                                          factor discussed previously. As explained in the table, nearly all
                                                          factors are assigned a number of  1, 2 or 3, and the factors are
                                                          summed to allow a comparison of all the alternatives. Environmen-
                                                          tal factors accounted for 5  of the 14 factors rated because air, sur-
                                                          face water and other media are each considered separately. Costs,
                                                          the last factor, are assigned ratings of 1,2,3,4 or 5 (rather than 1,2
                                                          or 3) because  costs are such an important  consideration and
                                                          because they vary over a large range.
                                                            A low score in the summed rating is best; i.e., it indicates that an
                                                          alternative is generally more feasible, has less  adverse environ-
                                                          mental impact and is more cost-effective than an alternative with a
                                                          higher score. The lowest-scoring alternatives are  (in  order) in-
                                                          creased  groundwater  monitoring,  landfilling  all of the  con-
                                                          taminated  soil  and landfilling only  the  hazardous soil. Because
                                                          landfilling all of the soil would cost about six times more than land-
                                                          filling only part of the soil, the option of landfilling all of the soil
                                                          was not considered further.
                                                          RECOMMENDATIONS

                                                            Two alternatives were considered in the final recommendations.
                                                          The first is to leave the arsenic in place and increase the ground-
                                                          water monitoring. This alternative is certainly the least expensive,
                                                          unless it is continued over a period of 10 or 12 years. In this case, it
474
SITE REMEDIATION

-------
                                                            Table 5
                                           Evaluation of Remedial Alternatives at Pcrham
Evaluation factors
1.
2.





3.
4.
5.
6.
7.
8.
9.
10.
Special engineering
considerations
Environmental impacts
Air
Surface water
Groundwater
Soil
Overall impact to
community
Operation, mainte-
nance, and monitoring
Time required for
implementation
Offsite disposal
needs and transpor-
tation plans
Legal constraints
Safety requirements
Ease with which it can
be phased into indi-
vidual operable units
Implementabi lity and
reliability
Total costs
Alternative 4b
Excavation, disposal, and landfarming
Minimal. Some design of landf arming operation is necessary.

Possible release of arsenic in fugitive dust and small amounts of arsenic gas during
excavation. Dust control measures such as light sprays of water on working face will
minimize dust. Personal protective equipment will reduce possible exposure to arsine.
Possible contamination from runoff during excavation. This can be eliminated by ex-
cavating during dry weather.
Existing groundwater contamination would remain, but would eventually (long-term)
diminish because source of arsenic would be removed.
Contaminated soil would be removed. Hazardous soil would be disposed of at an approved
landfill; nonhazardous soil would be landfarmed, resulting in small increases of
arsenic in the landfann plot.
Minimal negative impact from possible release of arsenic into the air during excava-
tion. Moderate positive impact due to removal of hazardous waste from area.
Operation would require excavation, soil sampling, and analysis to divide soil into
haurdous and nonhazardous fractions, transport to both a disposal facility and a
landfarm plot, waste application, and vegetation. Maintenance and monitoring are not
mandatory; however, some soil and soil-pore liquid monitoring is advisable.
Several weeks to months
Transport to an approved disposal facility could be by the facility or by a local
approved transporter of hazardous wastes. The nearest suitable landfill is 300 miles
away. Transport of nonhazardous soil could be by a local transporter.
Hazardous soil has to be accompanied by a manifest and taken to a RCRA-permitted
facility by a permitted transporter.
Personal protective equipment (Level C) is advised for excavation operations.
Units are excavation of hazardous soil, transport and disposal, excavation of nonhaz-
ardous soil, transport, and landfarming. Optional activities include landfarm site
revegetation, closure, and monitoring.
The only difficulty will be in adequately differentiating between hazardous and non-
hazardous soil.
$124,000
                           Table 6
                        Cost Summary
                Option
  1.  Increased groundwater monitoring
     (monitoring = $4,000/yr)

  ?.  Groundwater pumping and treatment

 3a.  Excavate and landfill all contaminated soil

 3b.  Excavate and landfill hazardous soil only

 4a.  Excavate and landfann all contaminated soil

 46.  Excavate all contaminated soil:  dispose'of
     hazardous soil, landfarm nonhazardous soil
                                           Approximate cost ($)
 15.000


361,000

398,000

 93.000

128.000

1Z4.000
is comparable to excavation and landfill disposal of only the hazar-
dous soil (which represents a one-time cost). At the present time,
the arsenic does not pose a risk because the clay cap prevents ex-
posure by air or surface water, and the contaminated groundwater
is not currently being used. Increased monitoring would make it
unlikely that a slug of contaminated groundwater could pass out of
the area without being  detected  in  time to  implement proper
remedial actions.
  The major disadvantage to this alternative, however,  is that it
does not remove  the  source of contamination or  treat  the con-
taminated groundwater. The immediate area of the trench could
never be used in any way that would disturb the plastic liner and
clay cap, and the  contaminated groundwater could not be used.
The community would still contain a hazardous waste site. The in-
creasing groundwater  monitoring program  without removal of
waste or groundwater would be perceived as a "do nothing" alter-
native, which would be socially and politically unacceptable.  The
option of leaving the waste in place and increasing the groundwater
monitoring, although  technically  acceptable because of minimal
risk of exposure,  is not socially acceptable and  is not  recom-
mended.
  The recommended remedial alternative at Perham is to conduct
extensive soil core sampling and analyses to define the boundaries
of the hazardous soil (that soil with an arsenic concentration of 500
mg/kg or more), and then to excavate the hazardous soil and trans-
port  it to  an approved  landfill  disposal  facility. The lesser-
contaminated soil (less than 500 mg/kg arsenic) would  remain in
place. The excavated area would be backfilled, and the plastic liner
and clay cap would be replaced. It is further recommended that the
ongoing groundwater  monitoring program be continued until the
levels of arsenic in all the wells remain below the  drinking water
standard. This alternative  is  the  second  most  cost-effective
measure. It  removes the source of most of the contamination, and
it minimizes transport and disposal costs by removing only that
                                                                                                 SITE REMEDIATION
                                                                              475

-------
                                               TRENCH  LIMITS
                                            ARSENIC CONCENTRATION

                                             FJ gOOmfl/ liter

                                             P^j 70-600 mg/IIU'

                                             r~]l-TO

                                             [  )l  mg/IIKr
                          FEET
                             Figure 6
     Vertical Soil Profile D-D1 with Arsenic Concentration Contours
portion of the soil that is most heavily contaminated.  Backfilling
the excavated area and replacing the plastic  liner and  clay  cap
would minimize further leaching of the arsenic remaining in place.
Because the groundwater would not be treated,  however, it must be
restricted  from  use. The community would most likely  be sup-
portive of this alternative.
Required Mapping
   The most difficult part of this alternative is the accurate mapping
of the area that needs to be excavated. This effort will require the
analysis of about  100 soil core samples for arsenic.  It  has  been
estimated (from earlier core sampling and analysis) that the area
with arsenic  concentrations exceeding 500 mg/kg is about 35  ft  
-------
                                           >1 ng/llter
                                           >500 i«)/lHer
                                        o  SOIL COKE SWUNG POIICTV
                                           J5 LOCATIONS EACH SA*>LED
                                           AT DEPTHS OF 4 ft, 6 ft,
                                           8 ft. AND 10 ft
                           Figure 7
              Proposed Soil Core Sampling Locations
plastic to prevent its contamination. The bucket of the backhoe
also should be lined with plastic to facilitate decontamination at the
end of the work effort.
Safety Procedures

  Persons hired to work in the hazardous area should be adequate-
ly trained in proper use of  protective equipment and  safety pro-
cedures. Medical examinations are recommended prior to site work
to ensure workers' fitness for this kind of work. Personal samplers
should be used on-site. According to the OSHA Inorganic Arsenic
Standard, CFR 29, Part 1910.1018, determinations of airborne ex-
posure levels shall be made from air samples representative of each
employee's exposure to inorganic arsenic over an 8 hr  period.
  As an initial precaution (pending results of air monitoring),  all
employees exposed to inorganic arsenic must be supplied with and
required to  wear a NIOSH-approved respirator and protective
suits, gloves and boots.  Level C protective equipment, modified to
close gaps at the neck or around the face,  is recommended. The
employer must provide workers with showers and clean changing
and eating facilities. The employer also must post the area where in-
organic arsenic is being handled.
  If the results of the initial  air monitoring indicate exposure levels
in excess of 5.0 /*g/m3,  personal protective equipment  (respirators
and  clothing) and showers  are required. Also required are con-
tinued medical surveillance,  recordkeeping and any other items in-
cluded in the site-specific health and safety plan. If the air levels are
less  than < 5.0  /tg/m3,  these precautions are not required,  but
careful handling of the arsenic is still necessary.
  At the beginning of the excavation, the clay caps and plastic liner
should be removed and retained. Excavation will proceed daily un-
til the project's completion. At  the  end of each day,  any con-
taminated disposal materials (respirator cartridges, clothing, etc.)
should be placed in the bucket of the backhoe and removed with
the plastic liner to the disposal trucks in which the hazardous soil
has  been  placed.  Waste  leaving the  site  by  truck  must be
manifested, with the-'State listed as generator.
  When excavation has been completed, clean fill should be placed
in the excavated area. After the area is leveled, the plastic liner and
clay  cap should be replaced. Thereafter, the clay cap should be in-
spected periodically (when the groundwater monitoring  wells are
sampled) to  ensure its integrity,  and repairs  should be  made as
needed.
Task
Contract wendMent
Soil core sampling
Soil analysis
Evaluation of results
Preparation of bid
documents
Hiring of contractors
Excavation
Transport and disposal
Closure
Reportl ng
July
—







August

—
—






September



—






October









November





__
__

De cectter







— -
                            Figure 8
        Implementation Schedule of Recommended Alternative


CONCLUSIONS

  This  recommended  alternative  could be  completed within  6
months. Because some landfills require a lead time of up  to  a
month to approve a waste, a sample  should be sent to landfills
under consideration as soon as possible. During this time, prepara-
tion of bid documents and selection of contractors for excavation,
hauling and disposal will be  started. The approximate schedule for
the required tasks is shown  in Figure 8. The schedule does not in-
clude continued groundwater monitoring which may entail several
years of periodic sampling and analysis.
                                                                                                SITE REMEDIATION
                                                           477

-------
     ENFORCEMENT  REMEDIAL  CLEANUP AT THE PETRO
           PROCESSORS  SITE; BATON  ROUGE,  LOUISIANA
                                           KEVIN G. GARRAHAN
                                        MICHAEL A. KILPATRICK
                                   U.S. Environmental Protection  Agency
                                   Office of Waste Programs  Enforcement
                                               Washington, D.C.
                                               DAVID  E. PRICE
                                   U.S. Environmental Protection  Agency
                                       Superfund Enforcement Section
                                                  Dallas, Texas
 INTRODUCTION

  The Comprehensive Response, Compensation and Liability Act
 (CERCLA) provided  the USEPA with  the authority to effect
 cleanup of hazardous waste sites through two mechanisms. The
 first is through initiation of federally financed  response actions
 funded  from  the $1.6 billion trust fund established under the
 statute.  The second, less publicized, mechanism is by activities in
 the USEPA's  Superfund Enforcement Program aimed at compel-
 ling private responsible parties to clean up sites thereby supple-
 menting limited Superfund  resources.  An example of USEPA's
 successful  application of  its  enforcement  authorities  under
 CERCLA  is the Petro Processors site near Baton Rouge, Loui-
 siana.

 Site Background

  The Petro Processors site is located  in Scotlandville, Louisiana,
 just outside of Baton Rouge. The site covers approximately 62
 acres and is comprised of two disposal areas, Scenic Highway and
 Brooklawn, located approximately I'/i miles apart. Both disposal
 areas are located on the banks of Bayou Baton Rouge  with the
 Brooklawn area located in the floodplain of the Mississippi River.
  The Scenic  Highway and  Brooklawn disposal areas were both
 owned by Petro Processors of Louisiana,  Inc., and operated from
 approximately 1964  to 1980. Waste generators  from  nearby
 petrochemical  and  chemical plants contracted with Petro  Pro-
 cessors for disposal of their waste products.
  The wastes  are present in  several  phases: solid,  semi-solid,
 gelatinous, sludge, liquid and mixed industrial debris (scrap metal,
 rubber, plastics). The main contaminants of concern are highly tox-
 ic volatile organic compounds present  in high concentrations.
  Both disposal areas  exhibit similar hydrogeologic features. The
 site is generally underlain by clay and silt deposits  with extensive
 seams of sand lenses. A shallow alluvial aquifer is present in a sand
strata located at depths of about 30 to 40  ft below mean sea level.
 Beneath the alluvial aquifer are several layers of clay and a major
aquifer commonly referred to as the  "400-foot sands aquifer."
This aquifer, which begins at depths of about 100 ft below mean
sea  level in  the vicinity  of the site, is an important water source for
industrial use and some private drinking wells in the area. Signifi-
cant levels  of  contamination have been  detected in the  alluvial
aquifer and in Bayou Baton Rouge. No contamination has yet been
detected in  the 400-ft deep sands aquifer in the vicinity of the Petro
 Processors  site.
  The Scenic Highway disposal area covers approximately 7 acres
 and is located  between a major highway, U. S. Route 61 (Scenic
                                                   Highway) and Bayou Baton Rouge. The site is a former borrow pit
                                                   from construction of Scenic Highway. Beginning in 1964, wastes
                                                   were disposed of in the borrow pit. The Scenic Highway disposal
                                                   area was filled and closed around  1974. Most of the liquid wastes
                                                   were solidified, fill dirt was added and plastic sheeting and a soil
                                                   cap were installed. The cap is now in disrepair.
                                                     A major concern at this site is the potential and actual exposure
                                                   and migration of toxic materials  by erosion from Bayou Baton
                                                   Rouge as well as subsurface seepage. It has been estimated that ap-
                                                   proximately 320,000 tons of waste material were disposed of at the
                                                   Scenic Highway disposal area.
                                                     Brooklawn is the larger disposal  area covering approximately 55
                                                   acres. It was opened in the late 1960s and was in full operation by
                                                   1970. It did not completely cease operation until a suit was filed by
                                                   the USEPA in July 1980. Approximately 19,000,000 gal of liquid
                                                   wastes and an additional 900,000  tons of non-liquid wastes were
                                                   disposed of at Brooklawn.
                                                     The Brooklawn disposal area includes two open waste disposal
                                                   pits,  a leveed cypress pond and several disposal pits which have
                                                   been filled and covered with soil. The Mississippi River periodically
                                                   inundates much of the area. As recently as June 1983, the cypress
                                                   pond was flooded and the water came within 4 in. of flowing over
                                                   the dike into the lower pit. A large disposal pit (now filled) was con-
                                                   structed over an  old  channel of  the bayou. This channel is a
                                                   suspected conduit for subsurface migration of wastes. Subsurface
                                                   contamination has been found outside of the disposal areas.
                                                   Enforcement History

                                                     The USEPA first became involved with the Petro Processors site
                                                   in May 1980 when inspections conducted by USEPA and the State
                                                   of Louisiana revealed that open waste pits were overfilled and in
                                                   danger of overflowing.  In addition, evidence showed that wastes
                                                   had been directly discharged from the  waste pits into Bayou Baton
                                                   Rouge.
                                                     The USEPA and  Louisiana filed a  lawsuit in Federal Court in
                                                   July 1980 seeking cessation of the discharges and a full investiga-
                                                   tion and cleanup of the site. The lawsuit named the owner/operator
                                                   and ten companies who had generated wastes sent to the Petro Pro-
                                                   cessors site as defendants. The lawsuit was filed under RCRA, the
                                                   Clean  Water Act and the Refuse Act and was later amended to in-
                                                   clude counts under CERCLA.
                                                     Subsequent to filing the lawsuit, the USEPA continued to in-
                                                   vestigate the  site to better define  the potential  threats to  public
                                                   health and the environment.  Those investigations revealed both
                                                   organic phase and aqueous phase contamination in the alluvial
478
SITE REMEDIATION

-------
aquifer beneath and migrating from the site. In addition,  con-
tamination was found in the sediments of Bayou Baton Rouge and
in other off-site areas. Site geologic conditions gave rise to serious
concern over the potential for contamination of the deeper (400-ft)
sands aquifer. The USEPA was similarly concerned about the con-
tinuing seepage of wastes from pits adjourning Bayou Baton Rouge
and the possibility of massive releases  in the event of failure of
deteriorating dikes around the open waste disposal pits.
  In response to the concerns expressed by the USEPA and the
State regarding the conditions of the open waste pits, the industry
defendants  entered  into  a stipulation,  filed with the  Court in
December 1981, whereby the defendants provided $218,000 to the
State in order to implement certain interim remedial measures.
Those measures, which were carried out by the State's contractor in
April 1982,  included reinforcing dikes, diverting surface water, im-
proving freeboard and providing better site security. When heavy
rains again caused liquid levels in the waste pits to rise dangerously
high in February 1983, industry defendants conducted an emergen-
cy removal  of aqueous liquids for disposal off-site in an injection
well, all under the  supervision of the USEPA and State on-scene
coordinators.
  After more than three years of technical discussions and negotia-
tions failed  to produce an agreement for remedial cleanup, trial in
the lawsuit was convened on Dec. 5, 1983; however, with extensive
participation by the  Court,  a  settlement agreement was reached
with the  industry defendants on that same day.  The settlement
agreement was embodied in  a  Consent Decree which  was lodged
with the Court Dec. 15,  1983.
  Under terms of the Consent Decree, the 10 industry defendants
agreed conceptually to a cleanup program for the entire Brooklawn
and Scenic  Highway disposal  areas.  Specifically,  the Consent
Decree requires the industry defendants to: (1) perform detailed
remedial planning activities;  (2) design,  construct and implement
the agreed upon conceptual closure plan which includes a state-of-
the-art landfill and groundwater recovery systems; and (3) maintain
perpetual operation, maintenance and monitoring of the site.
  If, after implementation of the closure plan, the monitoring pro-
gram detects a release or threat of release of contaminants from the
site, the industry defendants are required by the Consent Decree to
submit a supplemental remedial action plan to prevent or mitigate
any such release and to implement any additional remedial action.
Further, notwithstanding results of the  monitoring  program, the
government parties reserved their rights to reopen the case and seek
additional remedial action which may be necessary to prevent or
mitigate contamination in the deep sands aquifer in the event that
there may be changes in either land use near the site, aquifer use,
groundwater flow  direction  or velocity, or an endangerment is
presented by site conditions.
  In addition  to implementation of the closure plan,  which is
estimated to cost approximately $50 million and will take several
years to implement, the industry defendants  reimbursed the Hazar-
dous  Substance Response Trust  Fund  $600,000 to cover the
USEPA's investigation and enforcement costs. These funds will be
used for future site cleanups.
 CONCEPTUAL CLOSURE PLAN
  The conceptual closure plan attached to the Consent Decree pro-
 vides the technical framework,  details and design criteria for the
 agreed upon remedial action as well as the operation, maintenance
 and monitoring programs. Prior to design and implementation of
 the closure plan, the industry defendants will also perform several
 remedial planning activities.
  Included among  the remedial planning activities are:  a  site
 characterization  study  which will be essentially a  remedial in-
 vestigation and feasibility study  to determine the nature and extent
 of groundwater  contamination and to evaluate remedial alter-
 natives necessary to prevent or mitigate contamination of the deep
 sand aquifer; a health and safety plan; a quality assurance/quality
 control  program; a waste solidification testing program; a liner
compatibility testing program; and a study evaluating the suitabili-
ty of.proposed landfill sites.

Waste Removal

  The conceptual closure plan requires solid and semi-solid waste
materials from both the Scenic Highway and Brooklawn disposal
areas as well as contaminated sediments and soils from the cypress
pond, Bayou Baton Rouge and adjacent property to be  excavated
and mixed with appropriate solidification agents prior to disposal
in a  state-of-the-art landfill. The appropriate solidification agents
will be determined through the testing program  in the remedial
planning activities.
  Wastes to be removed from the Brooklawn and Scenic Highway
disposal areas will include all visible wastes plus 1 ft of soil or sedi-
ment from all faces of the excavation. On a portion  of adjacent
property which had been contaminated from a spill several years
earlier, three feet of soil and sediments will be removed even if there
are no visible wastes present. All excavated materials will be han-
dled and treated as hazardous. The excavated areas of the site will be
partially backfilled and  graded to provide drainage of  surface
runoff.  A protective cap system will be constructed and  include
three feet  of low  permeability clay  (10 ~7  cm/sec  or less),  a
vegetative layer and will  be seeded or hydromulched to  provide a
grass cover.

Geotechnical Tests
   Geotechnical testing will be performed in the Brooklawn bluff
areas to determine the feasibility of allowing in situ containment of
wastes buried in that area. Soil borings will be taken to  determine
the  existence of a continuous layer  of low permeability  clay
(aquitard) of  sufficient thickness to serve as a barrier to  vertical
waste migration. If an adequate clay layer does exist, then the bluff
disposal area will be encircled with cutoff walls keyed into the clay
with a multi-layer cap system installed over the disposal  area. Any
liquid wastes within the containment area will be either solidified
and  disposed, incinerated on-site, or treated, stored or disposed of
as hazardous waste in a RCRA permitted off-site facility. If an ade-
quate clay layer does not exist, then the wastes will be excavated,
solidified and disposed of in the state-of-the-art landfill.

Groundwater Treatment
   Two groundwater recovery pumping systems will be installed and
operated to extract and treat subsurface liquid contaminants.  The
first groundwater  recovery  pumping system will be  designed to
remove and treat any organic liquid phases present beneath the site.
The organic phase recovery wells will be installed  and operated in
those areas where organic phase liquids have been detected during
investigation  activities.  Organic  phase  recovery  wells  will be
pumped until the  organic phase liquids are no longer present as
determined by monitoring of effluent from the pump discharge.
   The second groundwater recovery system will be designed to ex-
tract and treat aqueous phase contaminants (contaminants dissolv-
ed in water) from the groundwater. This system  will be installed
where  appropriate at  both disposal areas to recover  and treat
aqueous phase  contaminants from the  shallow  alluvial  aquifer
located in sand lenses at  about elevation minus 40 mean sea level.
The aqueous phase groundwater recovery system will be pumped as
necessary (for a minimum of three months) to prevent contamina-
tion of the deeper uncontaminated deep sands  aquifer.  If  con-
tamination  cannot be prevented, additional  remedies may be
sought  to  prevent any  endangerment to  potential groundwater
users.
   Contaminated water from the waste  disposal pits, cypress pond
and  the well recovery system will be treated and discharged in ac-
cordance with an  NPDES permit. Organic phase  contamination
recovered from well recovery systems will be stabilized, solidified
and  disposed in the landfill, incinerated on-site, or treated, stored
or disposed of as a hazardous waste at an off-site RCRA permitted
facility.
                                                                                                SITE REMEDIATION
                                                           479

-------
State-of-the-Art Landfill

   A state-of-the-art landfill will be constructed on-site or on adja-
cent property. The landfill will meet the substantive technical re-
quirements of RCRA, although a permit will not be required.
   The landfill will incorporate a liner syster composed  of 3 ft of
clay with a permeability less than 10 -'  cm/sec, a flexible mem-
brane (synthetic)  liner and a leachate collection system. The ma-
terial and  thickness of the flexible membrane liner will be deter-
mined through compatibility testing conducted as part  of the re-
medial planning activities. The bottom of the liner system will be
above the  100-year  flood  elevation and the historic high ground
water elevation.
   A multi-layer  cap system which minimizes the infiltration of
liquids into the landfill in the long-term  will cover the top and sides
of the landfill. The cap system will  include a barrier layer com-
posed of both a layer of clay with a permeability of less  than 10 - '
cm/sec and a flexible membrane liner,  a porous drainage layer (if
necessary)  and a  vegetable layer to prevent erosion. The landfill
will be designed and operated to prevent run-on  from the peak dis-
charge of at least the 25-year storm and to collect and control run-
off from at least the 24-hr, 25-year storm event. Furthermore,  the
design  will employ appropriate  design  or  operational controls
necessary to prevent wind dispersal of paniculate matter.
   The landfill design will include  the location  and specifications
for groundwater  monitoring wells capable of detecting any poten-
tial releases of contaminants from the landfill into groundwater. If
                                                          the final landfill design includes a leachate detection system and
                                                          backup flexible membrane liner in addition to the required leach-
                                                          ate collection system, flexible membrane liner and clay liner, then
                                                          groundwater detection wells may not be required. However, in the
                                                          event that leachate is discovered in the leachate detection system,
                                                          a groundwater monitoring system will be installed.
                                                          REFERENCES
                                                           1. U.S. District Court, Middle District of Louisiana, Civil Action No.
                                                             80-358-B, "Consent Decree", United Slates versus Petro Processors
                                                             of Louisiana, Inc. el a/.. Feb. 1, 1984.
                                                          2. U.S. Environmental  Protection  Agency, "Action Memorandum—
                                                             Petro Processors Inc.," Aug. 1983.
                                                          3. CH2M-HJ1I, "Alternative Evaluation of Closure Plans—Petro Pro-
                                                             cessors of Louisiana, Inc.," July 1983.
                                                          4. CH2M-Hill, "Technical  Evaluation of the Proposed Closure Plans
                                                             for Petro-Processors, Inc.," May 1983.
                                                          5. CH2M-Hill, "Representation and Technical Evaluation of Data from
                                                             the  Petro  Processors  of Louisiana,  Fnc.  Uncontrolled Hazardous
                                                             Waste Sites".
                                                          6. TERA Corporation "Surface Characterization," Jan. 1983.
                                                          7. U.S. Army Corps of Engineers, Letter Report to USEPA, May 1983.
                                                          8. Geraghty & Miller, Inc., "Preliminary Hydrogeologic and  Ground-
                                                             Water Assessment  at  Petro-Processors, Inc.  Sites,  Baton Rouge,
                                                             Louisiana", Nov. 1981.
480
SITE REMEDIATION

-------
             SUBSURFACE GEOPHYSICAL INVESTIGATION
                                   AND SITE MITIGATION

                                         LEE TAYLOR LAWRENCE
                                      Chemical Waste Management, Inc.
                                   Environmental Remedial Action Division
                                                Oak Brook, Illinois
INTRODUCTION

  During 1977,  the  Illinois  Environmental  Protection Agency
(IEPA) obtained a circuit court order requiring the owner of a sand
and gravel quarry located in rural northern Illinois to remove
approximately 400 buried drums of paint wastes and sludges. The
wastes were allegedly buried at the operating quarry throughout the
60s and early 70s. It had been reported that portions of the quarry
were used to dispose of residential and industrial refuse as well.
  In 1979, ownership of the 170 acre quarry was transferred to the
County Conservation Department through a purchase arrange-
ment with the previous owner. The County's intent in acquiring the
sand and gravel pit was to develop  the entire area into a wilderness
preservation park complete with camping and hiking facilities and a
stream fed lake for fishing and boating. The lake would be created
by allowing the quarry to fill naturally with fresh water.
  The restoration plans and activities would be scheduled over a
period of ten years, while the economic value  of the quarry could
still be realized through continued mining operations. The ongoing
sand and gravel activities conveniently assisted preparation of the
site for the future park. The restoration proceeded smoothly until
August, 1983.
  During the development of a steeply banked area, which would
someday be the north shore of the proposed lake, excavation equip-
ment unearthed three rusted drum carcasses. At this point, the
restoration effort was stopped and County officials were notified.

BACKGROUND
  The County's initial response to the findings at the quarry was to
collect samples of material clinging to the  drum carcasses and have
the material analyzed by a state-operated laboratory. The labor-
atory reported the presence of aromatic  hydrocarbons similar to
the drummed wastes which had been exhumed by the previous site
owner/operator seven years earlier. County personnel scanned the
area with a magnatometer and confirmed their initial concerns
about the presence of many more buried metal objects  than had
originally been unearthed. An emergency meeting of the County
Commissioners was called by the president of the Conservation
District in order to review the findings and map out a strategy to
address the situation which was drawing increasing public concern.
  The  Board of Commissioners unanimously decided to conduct
whatever measures were necessary to remove and properly dispose
of the waste materials. A list of available remedial contractors was
obtained from the IEPA. Within three days of the Board of Com-
missioners vote to pursue cleanup  of the  site,  a contractors meet-
ing was held with potential remedial action firms.
  The outcome of discussions with potential contractors raised
several issues which the Commissioners needed to consider prior to
initiation of any on-site remedial activities. They were:
•Could cleanup costs be recovered through any state and/or fed-
 eral programs or would litigation be required?
•Once the cleanup was initiated,  how would the county know the
 extent of waste removed required?
•If materials  identified by way of the county's magnatometer sur-
 vey were excavated,  what  assurance would there be that other
 buried wastes were not present in the surrounding vicinity?
•Would there have to be special state and/or federal requirements
 for conducting the cleanup?
•Once the wastes were removed, where could they be properly dis-
 posed?
SITE INVESTIGATION

  In mid-August, 1983,  the  county  contracted with Chemical
Waste  Management,  Inc.'s  Environmental Remedial  Action
Division (CWM-ENRAC) for  implementation of remedial site in-
vestigations. The recommended initial  site investigations consisted
of the following geophysical procedures:
•Land surveying
•Terrain conductivity survey
•Ground penetrating radar survey
•Test excavations
•Multiple soil borings
  Due to the unknown extent of buried materials, a subsurface in-
vestigation consisting of the aforementioned tasks was advisable.
The objective  of these tests was to develop data for a remedial
action plan.
  In view of the drum removal activities seven years earlier, inves-
tigation was to cover 4.6 acres, thus providing geophysical data for
the area surrounding the exposed drum carcasses as well as check-
ing other potential burial locations. Before any on-site activity was
conducted, the entire area was monitored with a Century Organic
Vapor Analyzer (OVA). This survey determined that the initial site
investigations  could be safely  conducted in USEPA Level D pro-
tection, provided the  activities did not disturb the soil (Fig. 1).
Environmental air monitoring  was continued throughout all of the
site investigation phase.

Land Surveying

  The limits of the land area to be investigated were predicated on
the two exposed waste locations. The resultant area was a square
                                                                                         SITE REMEDIATION
                                                      481

-------
with approximate dimensions of 450 ft per side. Once the investi-
gatory zone had been delineated, a surveying crew established a
grid pattern and set grid stakes at 100 ft centers over the parcel and
at 25 and SO ft centers over those portions of the grid which
physically exhibited  the greatest potential  of being areas where
burial activities could have taken place. The grid pattern and asso-
ciated coordinate identifiers were keyed to known permanent fea-
tures in the immediate vicinity for possible future reference.

Terrain Conductivity Survey
  The terrain conductivity survey utilized a Oeonics Limited Model
EM31 conductivity meter. This meter generates a primary magnetic
field which induces electrical currents in the subsurface; the resul-
tant secondary magnetic Held is measured and evaluated by con-
ductivity measurement.  Terrain conductivity  is dependent  on the
porosity,  moisture content,  concentration of dissolved electro-
lytes, temperature and phase state of the moisture,  amount and
composition of colloids of both soil and rock and presence of in-
troduced high or low conductivity material.
  The terrain conductivity survey identifies conductivity anomalies
without regard to the source of the anomaly. As a result, the source
of the observed anomaly cannot be assigned based solely  on the
geophysical survey, but  must  be supported by material exposed at
the surface; i.e., historical  records, subsurface sampling or exca-
vation and direct observation. The magnitude of  conductivity
values is not significant in a survey intended to identify anomalies.
During this survey,  conductivity value adjustments were deter-
mined to assure that readings  on the instrument's display dial were
within a given range of the dial for background values. The survey
was conducted by measuring conductivity values on a 25 ft spac-
ing over the entire parcel and on a 12.5 ft spacing in the grid area
where most of the conductivity anomalies were found at 25 ft spac-
ing. Measurements were made at locations of grid stakes in the grid
area, at locations determined by pacing between  stakes where the
stakes were either 50 or 100  ft apart and in  the area of conduc-
tivity anomalies. The location accuracy was judged to be within 2
ft where pacing was used.
  The Model EM31  conductivity meter has its dipoles fixed 12 ft
apart. At  each point surveyed, the long axis of the instrument was
oriented in an east-west direction and conductivity values were
measured using two different equipment configurations, the hori-
zontal and vertical dipole mode. In the horizontal mode, approx-
imately 75% of the observed conductivity is estimated to be due to
material in the region from 0 to 10 ft below ground surface. In the
vertical mode, approximately  50% of the conductivity observation
is estimated to be due to material in the region from 0 to 10 ft below
ground surface, the majority of that response coining from below a
depth of 5 ft with little near surface influence. The remaining 50%
of the observed conductivity is estimated  to be due to material in
the region 10 to 20 ft  deep. In  this discussion, the horizontal
dipole mode is called the 10 ft mode and the vertical dipole mode is
called the 20 ft mode.
  Location and conductivity value information were noted on data
forms and then recorded on plan view maps  of the grid pattern.
Contours  of approximately equal conductivity values were  then
drawn. Conductivity values were transferred to a map of the grid,
and approximate contours of anomalous conductivity values were
drawn. From the conditions delineated through  the  conductivity
survey, a general site assessment was made with respect to the spe-
cific areas within the overall grid matrix which would be the focus
of additional geophysical investigations.

Ground Penetrating Radar Survey
  The ground-penetrating radar equipment  used  was an SIR Sys-
tem 8 manufactured by Geophysical  Survey  Systems, Inc.  The
system consists of a control  unit,  transducer (radar transmitter,
receiver and antenna), a magnetic tape recorder and a graphic chart
recorder. The instrument operates on 12 v DC obtained  from the
electrical system of the support vehicle used for data collection.
                                                                                   Figure I
                                                          Air Monitoring was Conducted at the Initiation of Site Activities and
                                                                          Throughout the Entire Project

                                                          Radar transducers operating at different frequencies yield greater
                                                        depth penetration of the radar signal, while higher frequencies, al-
                                                        though not able to penetrate the earth as deeply, give greater reso-
                                                        lution.  This  greater resolution gives the higher frequency trans-
                                                        ducer the  ability  to discriminate  between closely placed objects
                                                        and interfaces. The radar signal judged best was a 300 megahertz
                                                        transducer. This transducer  yielded  good near surface resolution
                                                        while still providing adequate depth penetration.
                                                          In operation, a brief pulse of electromagnetic energy is directed
                                                        into the subsurface. While this energy encounters an interface be-
                                                        tween two materials of different dielectric properties, a portion of
                                                        the energy is reflected back to the transducer. The reflected energy
                                                        is received by the transducer and processed within the control unit
                                                        where it is  amplified and the time differential between initial trans-
                                                        mission of the electromagnetic pulse and the  reception of the re-
                                                        flected  wave is determined.
                                                          The electromagnetic wave travels through the medium at a veloc-
                                                        ity dependent upon its dielectric characteristics, so the time differ-
                                                        ential can be converted into depth. The depth versus  time relation-
                                                        ship  can be established from a knowledge of the dielectric  con-
                                                        stant of the medium or, more commonly, from on-site determina-
                                                        tion of the depth of a visible radar target  (soil borings). No vertical
                                                        calibrations were  conducted for  this survey.  Approximate depth
                                                        determinations were made  based on generalized soil parameter
                                                        values.  The total depth shown on the strip charts was approximate-
                                                        ly 10 ft and the detected anomalies (discussed subsequently)  were
                                                        within the upper 3 to 4 ft.
482
SITE REMEDIATION

-------
  The electromagnetic pulse is repeated at a rate of 50 kilohertz
(50 x 101 cycles per second). The data are sent to the chart recorder
where a continuous record of the data is produced as the trans-
ducer is moved along the surface; data are also sent to the magnetic
tape recorder where the individual return wave forms are recorded.
  At the control unit,  the  operator  has an oscilloscope display
unit which continuously monitors the reflected wave form.  The
operator also has controls available to adjust and optimize the wave
form to produce the best output on the graphic chart recorder and
magnetic tape.
  As with any geophysical application, the results obtained from
the equipment depend  on site-specific  conditions. Physical con-
ditions which affect the strength of the return signal include the
amount of clay and moisture in the subsurface. In general, the site
was dry, but some pockets of moisture were observed on the sur-
face and those areas were detected by the radar. These higher mois-
ture zones did not appear to significantly affect the  system per-
formance nor did they interfere with the overall results; however,
the presence of clay did reduce the depth of penetration.
  The GPR provides information only along the line of antenna
tow for a width of approximately 2 ft. The antenna was manually
towed along grid lines using conductive cables which connected the
antenna to the control unit  located in a support  vehicle parked
nearby. The operator provided real time interpretation of the data
with the results of the GPR survey printed on a continuous  roll
paper copy. As signals indicating foreign bodies were observed on
the paper copy, the operator in the van called out and a technician
marked the suspect spot on the ground with spray paint.

DISCUSSION OF RESULTS
Terrain Conductivity Survey

  A total of 804 terrain conductivity measurements were made at
419 locations over the gridded 4.6 acre area.  Consistent back-
ground conductivity values were measured in the range of 70 to 90
millimho/m for the 10 ft mode and in the range of 75 to 95 mill-
imho/m for the 20 ft mode.
  The locations of the 100 millimho/m contours for 10 and 20 ft
modes approximately overlap. The 400 millimho/m contour for the
10 ft mode contained most of the same contour for the 20 ft mode
and is approximately twice the area. The 400 millimho/m contour
for the 10 ft mode contained most of the area where exposed drums
were observed.
  The terrain conductivity  survey  showed  strong  conductivity
anomalies at the known location of exposed drums. On this basis, it
was judged probable that there were other buried objects in the
region contained within the  400 millimho/m contour for the 10 ft
mode.
  The shapes of the 100 and 400 millimho/m contours  for the 10 ft
mode were similar.  For the area between these contours, it was not
clear if there were weaker sources of high conductivity present or if
the zone of influence of strong sources (within the 400 millimho/m
contour) extended to the 100 millimho/m contour. The similarity
of shapes of  the contours suggested that the latter might  be the
case. The sources of conductivity anomalies were probably present
at a depth of 5 to 20 ft.
  Due to soil characteristics in relation to the Geonics Limited
Model EM31 conductivity meter, nothing could be stated about the
presence of conductivity anomalies below a depth of 20 ft.

Ground Penetrating Radar Survey

  The approximate total length of the GPR survey was  1200 ft. The
GPR  signals indicating  the presence of a dielectric anomaly were
categorized as weak or strong and as having signatures typical of a
buried metal object or not.
  In the grid area, 38 dielectric anomalies were identified; ten were
categorized as weak signals, nine as strong signals, five as weak sig-
nals with signatures typical of buried metal objects and 14 as strong
signals with signatures typical of buried metal  objects.  The loca-
tions of these dielectric anomalies were recorded on a plot map.
  There were no vertical in situ controls (known depths of buried
targets) available at the time of the GPR survey. As a result, the
depths of the  strong and weak signals could not be determined.
Using typical soil properties, it was estimated  that all  identified
dielectric anomalies were at a depth of 4 ft or less. Because of the
presence of water and clay, the maximum depth of radar penetra-
tion reflected back to the antenna was estimated to be 10 ft or less.
  Due to the operating characteristics of GPR systems, it was not
possible to determine if there were any other dielectric anomalies
(e.g., buried  metal objects)  located  directly  below  where  an
anomaly was encountered.
  The locations of conductivity and dielectric  anomalies  were
shown superposed on a plot map. In the grid area,  12 of the 14
strong signals with signatures typical of buried metal objects were
located within the contours  of conductivity anomalies; 4 of the 9
strong signals  were within the contours; 2  of the 5 weak signals
with signatures typical of buried metal objects were within the con-
tours; and 1 of the 10 weak signals was within the contours.

GEOPHYSICAL SURVEY CONCLUSIONS
  Only two  relatively small  regions of anomalously high conduc-
tivity were found in the grid area. In one of these two areas, high
conductivity was found in both the 10 and 20 ft modes. In the other
area, one high reading was measured in  the 10 ft mode only. The
area around known locations of exposed buried drums was, for the
most part, within the 10  ft  mode, a very  high  conductivity con-
tour.
  In the grid area, the majority of the strong radar signals, includ-
ing almost all of those  with signatures typical of buried metal
objects, occurred  at locations within the contours of conductivity
anomalies. Almost all the weak signals are outside these contours.
Based on the results of the terrain conductivity  and ground pene-
trating radar surveys, specific regions within the overall grid matrix
were identified for confirmation as sources containing buried waste
by way of test excavations.
                           Figure 2
   Test Excavations were Conducted in Areas Which Exhibited a High
              Probability of Containing Buried Wastes
                                                                                                SITE REMEDIATION       483

-------
Test Excavations
  Areas to be initially excavated were identified both on a map as
well as in the field.  A crew employing USEPA Level C protective
equipment utilized  a  1 Vi  yd1 tracked excavator to  explore areas
where buried wastes were expected based on the results of the geo-
physical  surveys (Fig. 2).  Without exception, all areas which ex-
hibited  strong  anomalous characteristics  were confirmed  to  be
sources of buried drums.
  The excavation gradually penetrated the soil with trenches 2 to 4
ft wide and 15 to 20 ft long until: (1) buried drums or visible con-
tamination were observed, or (2) the test excavation trench reached
a depth of approximately 15 ft where clay strata were encountered.
All anomalous conditions were excavated  using  this trenching
method.
  During the excavation,  it was apparent that the sources of con-
tamination were drums buried in an area of the following dimen-
sions: 100 ft from north to south and 50 ft from east to west at an
average depth of approximately  13 to 18 ft. It was clearly evident
that the drums had  been haphazardly dumped into the hole as the
overall integrity of the containers had decayed so that the primary
waste materials encountered were distorted drum carcasses, visibly
contaminated sub-soils and miscellaneous pint and quart size paint
cans. Standing water, which was apparent approximately 4 ft below
the surface, engulfed the entire 50 ft x 100 ft burial site.
   During the initial excavation phase, field  technicians wearing
USEPA Level  B  protective equipment  obtained samples of drum
residues, surrounding soil and groundwater for analysis (Fig. 3).
   Samples obtained  during  the test excavations contained  the
following compounds:
Methanol             Toluene
2-Ethoxyethanol       3-( 1,1 -Dimethylethyl) Phenol
Ethylbenzene         Naphthalene
Xylene               Benzene
Propylene Glycol      Phenol
   Fifteen test excavation  pits were dug on the site,  clearly estab-
lishing the physical  parameters of the burial trench. Approximate-
ly 65 crushed and empty drum carcasses could be identified; how-
ever no materials were removed from any test pits once waste was
encountered. Therefore, the exact number of buried drums could
not be ascertained;  only the boundary limits of the disposal area
were established.
   The remedial investigation provided the working information
required to begin development of a remedial action plan to remove
wastes from the site; however, the limits of soil contamination out-
side the identified burial trench needed to be established due to the
potential of contaminant transport through the soil.
                           Figure 3
      Technicians Secured Residue Samples for Analytical Testing
                       and Identification
                                                        Multiple Soil Boringi
                                                          To determine whether there had been migration of buried wastei
                                                        into the surrounding soils, 44 holes were drilled radially around the
                                                        burial trench.  The outer ring of the boring was 150 ft from the
                                                        corner points of the burial trench. The holes were drilled using a
                                                        drill rig mounted on an all-terrain vehicle. Soil samples were ob-
                                                        tained using either a split spoon or an auger. The sampling spoon
                                                        and/or auger were decontaminated after each boring was made.
                                                          In general, each of the soil borings revealed a surface layer of
                                                        granular materials  consisting  primarily  of sand and gravel, at
                                                        depths ranging from 3.0 ft below present grade to 8.0 ft below
                                                        grade.  Each of  the soil borings  drilled  in this exploration  was
                                                        terminated in soft to hard silty clays, the majority of which were
                                                        from 10 to 30 ft  thick.  From the field boring data, it was deter-
                                                        mined that the clay layer was a continuous strata existing across the
                                                        limits of the exploration.
                                                          Interpretation of water activity below the site was accomplished
                                                        through real time measurement of water levels in the 44 bore holes
                                                        just after development and  24 hr later. Although this did not pro-
                                                        vide the exacting data which could be generated through piezo-
                                                        meter studies,  the relative relationship between the water  level in
                                                        the different holes showed  that the variance  in water level was
                                                        only ±0.2 ft across the entire area where holes were drilled. This
                                                        minor variation, given the  protocol employed,  resulted in a con-
                                                        clusion that the water associated with the burial trench had a high
                                                        probability of being perched or at least had a very low rate of trans-
                                                        port; this conclusion  is plausible given the underlying soil con-
                                                        ditions.  The mere presence of the underlying water necessitated
                                                        that it be adequately addressed in  formulating the remedial action
                                                        approach for removal of the wastes.
                                                          Statistically selected samples from various depth intervals from
                                                        the 44 borings were analyzed to detect contaminants within and
                                                        outside  of the burial zone. There  was no contaminant transport
                                                        away from the source. Contaminants detected in samples of  soil
                                                        from within the burial trench were of a character similar to those
                                                        compounds identified during the test excavations.

                                                        REMEDIAL ACTION PLAN
                                                          The remedial investigation provided valuable data for develop-
                                                        ment of a remedial action plan. At the conclusion of the  explor-
                                                        atory activities, the following investigative objectives were met:
                                                        •Given the geophysical capabilities of the methods  employed dur-
                                                         ing the investigation, the source of buried wastes within the area
                                                         of concern was  isolated to a burial trench which had approx-
                                                         imate dimensions of 50 ft x 100 ft x 15 ft.
                                                        •The area which had buried wastes removed during an excavation
                                                         seven years earlier resulted in negative findings with respect to the
                                                         various geophysical surveys.
                                                        •Upon defining  the burial  trench, soils beyond the perimeter of
                                                         this trench tested free  of  contaminant transport  attributable to
                                                         the buried wastes.
                                                        •An identification of  water encircling the site was made during
                                                         drilling activities.
                                                        •Analysis  conducted on buried wastes provided the data necessary
                                                         to specify treatment/disposal alternatives.
                                                        •The volume of contaminated soils and drum carcasses was esti-
                                                         mated at 2,450 yd' ± 250 yd1.
                                                          Following a detailed analysis of the data developed during the in-
                                                        vestigation, a remedial cleanup approach  and  budgetary estimate
                                                        was formulated for review by the County Board of Commissioners.
                                                        The highlights of the remediation plan included:
                                                        •Complete removal of contaminated  solids (soils and drum car-
                                                         cases) from within the boundaries of the identified waste burial
                                                         trench  down to the interface of the underlying clay strata
                                                        •Containment of contaminant transport during remedial activities
                                                         through the development of a barrier wall to encircle the area to
                                                         be excavated
                                                        •Environmental  health  and safety protocols including a  contin-
                                                         gency program to be in effect during the cleanup effort
484
SITE REMEDIATION

-------
•Disposal of identified waste materials at a permitted secure chem-
 ical waste TSDF
•Specifications for the aforementioned remedial approach
  Upon recommendation by the County task force responsible for
evaluating the environmental  situation facing the Conservation
District, the Board of Commissioners voted to proceed with site
mitigation activities at the sand and gravel quarry. The Conserva-
tion District would  finance  100% of the cleanup  by employing
usury funds  generated  through  the commercialization of the
quarry; therefore, no tax monies or special funding provisions
would be required to remove the wastes.  Within five days from
the Board's authorization to proceed  with the cleanup, a remedial
action contractor had been selected with mobilization of personnel
and equipment initiated.
SITE MITIGATION

  Prior to actually removing the  identified waste materials from
the burial trench, it  was necessary for the on-site contractor to
construct a 360° barrier wall as a means of controlling potential
contaminant transport away from the trench.
  Numerous cut-off wall alternatives would meet  the objectives of
the project;  however, the selected approach would require rapid
installation while still providing the necessary restriction of poten-
tial contaminant transport. It was decided that, due to the schedule
required to exhume the wastes (six to  ten days) and the knowledge
that the burial trench area sat on top of a continuous clay layer, the
most efficient and cost-effective technique would be  a synthetic
barrier wall.
  The  synthetic  barrier wall  installation  was  straightforward.
Based on results of soil analyses conducted during the geophysical
surveys, a trench approximately 4 ft  wide was excavated around
the entire burial site. The trench placement was made at a distance
of 5 to 10 ft from the known boundaries of the burial trench and
to an average depth of 20 ft (approximately 2 ft into the underlying
clay strata).
  During excavation of the trench, which would  contain the syn-
thetic barrier wall, the installation field crew fabricated the 80 mil
high density polyethylene (HOPE) so it could be conveniently fitted
into the awaiting trench. Installation of the  HDPE liner necessi-
tated that standing water, which infiltrated the trench, be removed
prior to installation of the synthetic barrier. The liner was lowered
into the trench so that a "U" shaped envelope was created; this
envelope was keyed to the subsurface clay strata. Upon completion
of the liner placement, imported clay backfill was introduced into
the liners envelope, effectively cutting off the waste burial trench
from the surrounding uncontaminated areas.
  In addition to the barrier wall, a waste staging and loading pad
was constructed and lined with 80 mil HDPE, thus creating an ex-
clusion zone  which would be protected from cross contamination
of clean areas with the waste materials.
  Upon completion of site preparation, upwind and downwind air
monitoring stations were positioned. A personnel decontamination
facility, field command post and a vehicle decontamination  area
and tarping station were also established.
  Actual removal of the contaminated soils and drum carcasses re-
quired only seven working days, during which time 200 loads of
hazardous wastes  were removed and transported  to a  permitted
chemical waste disposal facility. In addition to the contaminated
solids encountered, it was necessary to remove several tanker loads
of potentially contaminated water which had infiltrated the exca-
vation trench. Once the wastes had been excavated and removed,
the barrier wall was left intact while the staging and loading pad
and associated materials were removed and disposed of off-site at a
permitted facility. Backfilling and grading of the excavation trench
with imported clay was accomplished subsequent to the site mitiga-
tion.

CONCLUSIONS
  The presence of uncontrolled hazardous wastes continues to
create environmental  concerns  throughout the United States.
Buried wastes present unique problem-solving situations, and this
project exhibits just one possible approach to the solution of an
environmental problem. As more and more abandoned hazardous
waste sites are  discovered, there will be an increasing need for a
developing industry to provide technologies and methodologies to
safely deal with these concerns. ENRAC is pleased to have worked
on this project and to have increased its firsthand knowledge in this
type of mitigation work as it strives to become a more experienced
and important factor in the emerging site mitigation segment of the
hazardous waste industry.

REFERENCES
1.  McNeill, J.D., Electromagnetic Terrain Conductivity Measurements at
   Low Induction Numbers, Technical Note TN-6, Geonics Ltd., 1980.
2. Geophysical Survey Systems, Inc.  Instrument Literature-SIR System 8,
   1982.
                                                                                                SITE REMEDIATION
                                                          485

-------
      GUIDELINES FOR DECONTAMINATING BUILDINGS,
    STRUCTURES AND  EQUIPMENT AT SUPERFUND  SITES
                                               M.P. ESPOSITO
                                                 J. McARDLE
                                                 J.S. GREBER
                                         PEDCo Environmental, Inc.
                                               Cincinnati, Ohio
                                                  R. CLARK
                                       Battelle Columbus Laboratories
                                               Columbus, Ohio
INTRODUCTION
  CERCLA established a dual-phase program for responding to
environmental problems caused by hazardous substances. The "re-
moval program" involves cleanup or other actions taken in re-
sponse to emergency conditions or on a short-term or temporary
basis. The "remedial program" involves response actions that tend
to be long-term in nature and permanently remedy problem sites.
  To be eligible for cleanup under Superfund, a site must be in-
cluded on the National Priorities List (NPL).  As of this writing,
406 sites appear on the NPL which was promulgated  by the
USEPA on Sept. 8,  1983. Currently, the USEPA is proposing the
addition of 133 new sites to the list.
  As the number of sites on the NPL grows and as removal and
remedial activities at Superfund sites accelerate, the task of decon-
taminating buildings, structures and construction equipment will
become increasingly important. These items often represent large
capital investments,  and the costs of dismantling and disposing of
such structures in a  secure landfill can be very expensive. The ob-
jective of an effective decontamination program, therefore, is to re-
turn contaminated buildings, structures and equipment to active,
productive status.
  This study had as its goal the development of a general guide for
government personnel, cleanup contractors  and other  individuals
responsible for planning and executing decontamination activities
at Superfund sites.
TECHNICAL APPROACH

  In the fall of 1983, a survey was made of ongoing decontamina-
tion activities at 50 Superfund sites across the country.  These sites
were thought to have potentially  contaminated buildings, struc-
tures and equipment, and this survey was conducted to gather in-
formation on: (1) the types of contaminants of most concern and
(2) the methods currently being proposed for use for decontamina-
tion of the buildings, structures and equipment in place at these
sites. Contractors and numerous other individuals with direct ex-
perience in both  Superfund and non-Superfund related programs
involving decontamination of dioxins, explosives, PCBs and other
toxic wastes from buildings and  equipment were contacted. In
addition, a thorough search of published literature for informa-
tion on decontamination methods  was conducted through com-
puterized search services.
  From  these surveys,  a decontamination  data  base  containing
state-of-the-art information on specific cleanup methods and their
applications, as well  as guidelines for developing site-specific clean-
up strategies, was developed.
                                                   RESULTS
                                                     The 1983 survey of building/equipment decontamination prac-
                                                   tices at Superfund sites revealed that the contaminants of most con-
                                                   cern at these sites included: asbestos, acids and alkalis, dioxins,
                                                   explosives, heavy metals, cyanides, low-level ionizing radiation,
                                                   organic solvents,  pesticides and PCBs.  The methods used to re-
                                                   move  these substances from buildings,  structures and equipment
                                                   are few in number and rarely documented in detail. For example,
                                                   it is common practice to steam clean equipment such as backhoes,
                                                   bulldozers and drilling augers, but testing to verify that the con-
                                                   taminants of concern  have been adequately removed is generally
                                                   not performed.
                                                     Contaminated buildings and structures are seldom cleaned and
                                                   returned to active use. More often, they are closed and barricaded
                                                   to prevent further entry and exposure until sometime in the future
                                                   when  a solution regarding [heir deposition can be found.  Some
                                                   buildings are  torn down and buried in landfills. Contaminated
                                                   underground structures such as tanks, sumps and sewers are some-
                                                   times filled in place with concrete to prevent their reuse.
                                                     Because these findings clearly pointed  to the need for basic guid-
                                                   ance material  on decontamination methods, the remainder of the
                                                   project was devoted to developing a manual or user's guide for this
                                                   purpose.  The handbook which  was developed  presents general
                                                   guidelines for developing a rationale and strategy (Fig. 1) for deal-
                                                   ing with the prospect of decontamination including guidance and
                                                   information for selecting the least-costly method(s) that are tech-
                                                   nologically feasible and that will effectively reduce contamination
                                                   to predetermined levels. Steps in the process  include: (1) deter-
                                                   mining the nature and extent of contamination; (2)  developing
                                                   and implementing a site-specific  decontamination  plan;  and (3)
                                                   evaluating decontamination effectiveness.
                                                   Step I
                                                   •Querying former employees, searching old business records, in-
                                                    spection reports and news stories
                                                   •Conducting a visual site inspection
                                                   •Collecting and analyzing samples from the contaminated surfaces
                                                    or structures.
                                                   Step 2
                                                   •Identifying the future intended use of buildings, structures and
                                                    equipment
                                                   •Establishing  decontamination target levels for the contaminants
                                                    present
                                                   •Identifying and evaluating potential decontamination methods
                                                   •Selecting the  most appropriate method(s) for achieving the de-
                                                    contamination target levels
486
SITE REMEDIATION

-------
                                ,	I
DETERMINE NATURE AND
EXTENT OF CONTAMINATION
 DEVELOP AND IMPLEMENT SITE-
 SPECIFIC DECONTAMINATION
 PLAN
 EVALUATE DECONTAMINATION
 EFFECTIVENESS
                                              O/JEP.Y row
                                              EMPLOIEES"
                                         SEARCH OLD BUSINESS RECORDS.
                                            INSPECTION REPORTS.
                                             AND NEWS STORIES
                                        CONDUCT VISUAL SITE INSPECTION
                                          COLLECT AND ANALY7E SAMPLES
                                         IDENTIFY FUTURE UTEHDED
                                         USE OF IU1LDINGS, STRUC-
                                         TURES. AND EQUIPMENT
                                         ESTABLISH DECOMAK:U.-ION
                                         TARGET LEVELS FOR COX1AK-
                                         1NANTS PRESENT
                                          IDENTIFY AND EVALUATE
                                          POTENTIAL DECON7«*:WTION
                                          METHODS
                                         SELECT HOST APPRCPF.UTE
                                         METHOiKS) FOR ACHIEVING
                                         DECON'AKINATlOn TARiET
                                         LEVELS
                                          DETERMINE UOSKER HEALTH
                                          AND SAFETY REQUIRiBEkTS
                                          WHITE SITE DECONT«:s*TION
                                          PLAN
                                          HIRE CONTRACTOR HO INITIATE
                                          CLEANUP
                                              REIHSPECT SITE
                                             COLLECT AND ANiLYZE
                                                 SAMPLES
                                           DETEWIIHE WHETHER TARGET
                                           LEVELS HAVE BEEN REACHED
                                         REPEAT OK MODIFY DECONTAMINATION
                                             PROCEDURE AS NECESSARY
                         Figure 1
 Flow Diagram Illustrating Sequence of Steps for Developing a
                 Decontamination Strategy
                                       •Determining  worker health and safety  requirements  (training,
                                         medical surveillance, personal protective equipment, site safety)
                                       •Writing the site decontamination plan
                                       •Estimating costs
                                       •Hiring the contractor and initiating cleanup
                                       Step 3
                                       •Reinspecting the site for evidence of residual contamination
                                       •Collecting and analyzing samples from the decontaminated area
                                       •Determining whether the target levels  for residual contamination
                                         have been reached
                                       •Repeating and, if necessary, modifying the decontamination pro-
                                         cedures until satisfactory results are obtained
                                          Descriptions of actual building decontamination efforts at both
                                       Superfund and non-Superfund sites are .included as case studies in
                                       the report.
                                          The manual contains process  descriptions for 21 decontamina-
                                       tion methodologies including both traditional and developmental
                                       techniques  (asbestos  abatement,  absorption,   demolition,  dis-
                                       mantling, dusting/vacuuming/wiping, encapsulation, gritblasting,
                                       hydroblasting/waterwashing,   painting/coating,    scarification,
                                       RadKleen,  solvent washing, steam cleaning, vapor-phase solvent
                                       extraction,  acid  etching, bleaching, flaming, drilling and spalling,
                                       K-20,  microbial degradation and photochemical  degradation).
                                       The usefulness of each cleanup method with various combinations
                                       of contaminants and materials is given.
                                          Potential cleanup methods are identified in a two-dimensional
                                       matrix (Table  1) matching contaminants (asbestos,  acids, alkalis,
                                       dioxins, explosives, heavy metals and cyanides, low-level radiation,
                                       organic solvents, pesticides,  PCBs and all contaminants in liquid
                                       form) with materials/surfaces (all wall, ceiling and floor surfaces;
                                       glass; plastic; metal; wood; brick; concrete; aluminum; and equip-
                                       ment and auxiliary structures).
                                          Finally,  the  manual  describes  safeguards  for  protecting  the
                                       health and safety  of decontamination workers during site opera-
                                       tions.  Topics  covered  include personnel training,  medical sur-
                                       veillance, personal protective equipment and site safety.

                                       CONCLUSIONS AND RECOMMENDATIONS
                                       FOR ADDITIONAL RESEARCH

                                          As a result of  this study, it has become clear that all future own-
                                       ers of decontaminated buildings and structures on Superfund sites
                                                                Table 1
                           Potential Decontamination Methods for Various Contaminant/Material Combinations9
Material/surface
All Mil, oiling, mil
floor surfaces
61. SI
Plastic
Hetil
Hood
Brick
Concrete
Equipment and military
structures (bulldozers,
electrical transformers
and capacitors, sumps,
ventilation ducts, etc.)
Contaminant
Asbestos
1.3.4.5
1.3,4.5
1,3,4,5
1,3.«,5
1.3.«,5
1.3.4.5
1,3,4,5
1.5
Acids
3,4,6.9.
16
3,4.6.9.16
3.4,6,9.16
7
7
7
7,10, It
4,6,11,12
13
Alkalis
3,4.6,9
3,4,6,9
15
7.15
7,15
7.15
7,10,15,16
4.6,11,12
13
Dioxins
3,4,5,6,9,
21
3,4,5,6,9.21
15
7.15
7,15
7.15
7. 10. 15. IB
4,5,6,11,
12,13,21
Explosives
3.4.6,9,13
6,16
B.15,16
7,8,15,16.
17
7,15,16,20
7.8,15,17,
20
7.8,10,15,
16,17,18.
20
4.6,8,11,
12.13.17
Heavy
•etals
and cyanide
3.4,5,6,8.
9,16
3,4,5,6.8,9,16
15
7.15
7.15
7,15
7,10,15,18
4,5,6,7,6,
11.12,13
Low-level
radiation
3.4.5.6.9
8
8,15
7.8,15
7.15
7,8,15
5,7,8,10,
15,17,16
4,6,8,11,
12,13
Organic
solvents
3,4,6,9
3,4,6,9
15
7,15
7,15
7,15
7,10,15.18
4,6,11.
12,13
Pesticides
3,4,6,9,13
16
15,16
7,15,16
7.15,16.20
7,15,20
7, 10, 15,14
18.20
4.6,11,
12,13
PCB's
3,4,5,6,9
3.4.5.6.9
15
7.15
7,15,19
7.15.H
7,10,15,18.
19
4,5,6,1:.
12,13,14
All con-
tamlMnts In
1 Iquid form
?
<
2
:
'
'
2
!
                                                             15.
Ke> for decontamination methods:
 1. Asbestos abatewent
 2. Absorption
 3. Demolition
 4. Dismantling
 5. Dustlng/vecuumlng/wlplng
 6. Encapsulation
Etch cell In the matrix represents • specific contaminant/substrate combination and contains numbers
corresponding to decontamination methods that either have been used In the specific Interactions, or
have the potential for such use. Used on available Information 1n the published literature.  Each
•ethodology can be used alone or 1n conjunction with one or more of the other procedures.
 7. Abrasive gritblasting
 B. Hydroblasting/waterwashing
 9. Painting/coating
10. Scarification
11. RadKleen
                                                             12.  Solvent washing
                                                             13.  Steam cleaning
                                                             14.  Vapor-phase solvent extraction
                                                                 Acid etching
                                                             16.  Bleaching
17.  Flailing
18.  Drilling and spalling
19.  K-20 coating
20.  Hlcroblal degradation
21.  Photochemical degr>
-------
should be  advised  of the nature of the contamination  that  was
present, the cleanup methods used and levels of any residual con-
taminants. Ensuring the transfer of such information from one site
owner to the next will require a method for permanently recording
this information. Regulations requiring the addition of such in-
formation  to the property deed, as is required in the deed of all
RCRA-permitted facilities, may be a workable solution.
  The handbook which was  developed  will provide much of the
guidance needed by  site cleanup personnel for decontaminating
buildings, structures and equipment. However, additional research
is needed to bridge gaps  in the state of the art in four key areas.
First, and  perhaps most importantly,  sampling methods for de-
termining the type  and degree of contamination existing on build-
ing/structure/equipment surfaces,  both before and after cleanup
efforts, are poorly developed, documented and  verified. Similarly,
subsurface sampling  techniques (such as corings) for determining
the depth  of contamination  in porous substances (such as con-
crete or wood floors)  have not been  adequately developed and
documented. Although "wipe tests" are often  referred  to, in site
records, the actual  methodology used is rarely described  in enough
detail to allow simulation or reproduction by others, and the tech-
nique itself  is known to be  inadequate for quantitatively  trans-
ferring contaminants from surfaces to wipes or swabs. Additional
research in this area is badly needed.
  Second, many of the decontamination  techniques described in
the manual were  developed specifically by the U.S. Army's Installa-
tion Restoration  Program. Their applicability to contaminant/ma-
terial combinations encountered at Superfund  sites has  not been
fully explored. Even where decontamination techniques are indi-
cated for certain  contaminant/material combinations, the more de-
tailed methodology descriptions should be consulted for any future
work that may be required before the methods are selected.
  Third, the effectiveness of  many decontamination methods cur-
rently in use has not been verified and documented. For example,
the degree to which  steam cleaning removes dioxin-contaminated
soil particles from drilling augers has not been established.  De-
contamination methods that  have not previously been applied to
specific contaminant/substrate combinations  but show a strong
potential applicability should be validated  in pilot investigations.
Additions/deletions  to the matrix  should  be  made accordingly.
New decontamination technologies that become  available also
should be evaluated and added to the matrix.
  Fourth,  a formal, systematic approach for determining accep-
table levels  of contaminants  remaining in and on building  and
equipment surfaces does  not currently exist. As a result, guidance
on the "How clean is clean?" issue and the establishment of target
levels could  not  be included in this manual and must continue to
be addressed on a case-by-case basis.
ACKNOWLEDGMENTS
  This work was funded by the USEPA, Municipal Environmental
Research Laboratory, Solid and Hazardous Waste Research Divis-
ion, Cincinnati, Ohio under Contract No. 68-03-3190. Ms. Naomi
P. Barkley was the Project Officer. Others contributing to the work
were Ann Crone  and Mike Hessling of PEDCo and Ann Langham,
Cindy McCandlish, Scott Brown and John B. Hallowell of Battelle.
SELECTED BIBLIOGRAPHY
Act II—How Clean is Clean? Hazardous Waste Report, 5(9), 1983, 13.
Battelle Columbus Laboratories, Final Report on Evaluation of Encapsu-
lants for Sprayed-on Asbestos-Containing Materials in Buildings.  Pre-
pared for the USEPA under Contract No. 68-03-2552, 1979.
Battelle  Columbus Laboratories, Final Report  on Evaluation  of Encap-
sulants for  Sprayed-On Asbestos-Containing  Materials  in  Buildings.
USEPA, Industrial Environmental Research Laboratory, Cincinnati, OH,
1979.
Crosby,  D.G. "Methods  of Photochemical  Degradation of Halogenated
Dioxins in View of Environmental Reclamation," In: Accidental Exposure
to Dioxins. Human Health Aspects, Academic Press,  Inc., New York, NY,
1983.
Dempsey, K.B., "Biotechnology Aids Disposal," Plants, Sites and Parks,
Sept./Oct. 1982, 1-8 +
                                                          Benecke, P., el al., Development of Novel Decontamination and fnerting
                                                          Techniques for Explosives-Contaminated Facilities,  Phase f—Identifica-
                                                          tion and Evaluation of Concepts.  Vols. 1 and 2. DRXTH-TE-CR-83alI,
                                                          July 1983.
                                                          Bromley, J., D.C. Wilson, and E.T. Smith, "Remedial Measures Follow-
                                                          ing Accidental Release of Dioxin," Chemosphere, 12,  1983, 687-703.
                                                          CH2M Hill. Hazardous Waste Site Investigation Training Course Guide.
                                                          Nov. 1983.
                                                          Di Domenico, A., el al., "Accidental Release  of 2,3,7,8-Teirachlorodi-
                                                          benzo-p-dioxin (TCDD) at Seveso, Italy. I. Sensitivity and Specificity of
                                                          Analytical Procedures Adopted for TCDD Analysis," Ecotoxicology and
                                                          Environ. Safely, 4. 1980, 283-297.
                                                          Di Domenico, A., el al. "Accidental Release  of 2,3,7,8-Tetrachlorodi-
                                                          bcnzo-p-dioxin (TCDD) at  Seveso, Italy:  HI.  Monitoring of  Residual
                                                          TCDD  Levels in  Reclaimed  Buildings,"  Ecotoxicology and  Environ.
                                                          Safety, 4, 1980,321-326.
                                                          Di Domenico, A., el al. "Accidental Release  of 2,3,7,8-Telrachlorodi-
                                                          benzol-p-dioxin (TCDD) at  Seveso: Assessment of  Environmental  Con-
                                                          tamination  and  of  Effectiveness  of Decontamination  Treatments,"
                                                          CODA TA Bulletin. 29, 1978. 53-59.
                                                          Furhman, D., Decontamination Operations at Gateway Army Ammu-
                                                          nition Plant. DRXTH-AS-CR-83250, Nov. 1983.
                                                          Ghezzi, I., "Potential 2,3,7,8-Tetrachlorodibenzo-p-dioxin  Exposure of
                                                          Seveso Decontamination Workers," Scandanavian J. of Work  Environ-
                                                          ment and Health, 5(Suppl. 1), 1982, 176-179.
                                                          Haincs, R.V. and  W.W. Kelley., Frankford Arsenal Decontamination/
                                                          Cleanup Operation—Cleanup and Demolition of the 400 Area,  Rockwell
                                                          International, Pub. No. N505TI000054, Nov. 1980.
                                                          Hawthorne, S.H.,  "Solvent Decontamination of PCB  Electrical Equip-
                                                          ment." In. IEEE Conference Proc., Vol. IA-18, July 1982.
                                                          Johnson, W.R., Frankford Arsenal Decontamination/Cleanup Operation
                                                          —Standing Operating Procedures for Cleanup of Heavy Metals and Ex-
                                                          plosive Residues and Radiological Contamination of Buildings at Frank-
                                                          ford Arsenal, Rockwell International, Pub. No. N5050P000009,  Apr.
                                                          1980.
                                                          Johnson, W.R., Frankford Arsenal Decontamination /Cleanup Operation
                                                          —Standing Operating Procedures for Cleanup of Heavy Metals and Ex-
                                                          plosive Residues From Buildings at Frankford  Arsenal, Rockwell Inter-
                                                          national. Pub. No. N5050P000014. Aug. 1980.
                                                          Jones, W.E.  Engineering and Development Support of General Decon
                                                          Technology for the U.S. Army's Installation Restoration Program. Task
                                                          5, Facility Decontamination,  Defense Technical Information Center, Alex-
                                                          andria, VA. Pub. No. 49-5002-0005. July 1982.
                                                          Marion, W.J., and Thomas, S., Decommissioning Handbook, DOE/EV/
                                                          10128-1. Nov. 1980.
                                                          Natale, A., and H. Levins.,  Asbestos Removal  and  Control. An Insiders
                                                          Guide to the Business, Sourcefinders, Voorhees, NJ, 1984.
                                                          New York State Office of General Services, The Binghamton Slate Office
                                                          Building Cleanup: A Progress Report Update. Albany, NY, Jan. 1983.
                                                          Noe, L., "Reclamation of the TCDD-Contaminated Seveso Area," In:
                                                          Accidental Exposure to Dioxins. Human Health  Aspects. Academic Press,
                                                          Inc..New York, NY, 1983.
                                                          Personal communication from V.G. Rose, Pacific Gas and Electric Com-
                                                          pany, San Francisco, CA. Apr. 17, 1984.
                                                          Pocchiari, F., "2,3,7,8-Tetrachlorodibenzo-p-dioxin Decontamination. In:
                                                          Chlorinated Phenoxy Acids  and Their Dioxins. Mode of Action, Health
                                                          Risks and Environmental Effects." Ecological Bulletin, 27, 1978,67-70.
                                                          Rockwell International, Final Report for the Frankford Arsenal Decon-
                                                          tamination/Cleanup Program, DRXTH-FE-CR-800,  Dec. 1980.
                                                          Roos, K.S., and P.A. Scofield, "Health and Safety Considerations: Super-
                                                          fund Hazardous Waste Sites," Proc. National Conference on  Manage-
                                                          ment of Uncontrolled Hazardous Waste Sites,  Washington. D.C., Oct.
                                                          1983,285-290.
                                                          State of Georgia, Department of Natural Resources, Environmental Pro-
                                                          tection Division, Luminous  Superfund Project  Report: Remedial Action
                                                          for the Removal of Ra-226 Contamination at  the  Luminous Processes,
                                                          Inc. Site in Clark County, GA, Aug. 1982.
                                                          Strahl, H., Frankford Arsenal Dtcontamination/Cleanup Operation-
                                                          Standing Operating Procedures for Cleanup of Explosive Residues From
                                                          Buildings  at  Frankford Arsenal.  Rockwell  International.  Pub. No.
                                                          N505P000018, Oct. 1980.
488
SITE REMEDIATION

-------
      CONTINGENCY  PLANNING FOR REMEDIAL  ACTIONS
                            AT  HAZARDOUS WASTE SITES

                                                 ROBERT GOLTZ
                                               M. SHAHEER ALVI
                                        SALVATORE BADALAMENTI
                                    U.S.  Environmental Protection Agency
                                                      Region II
                                               New York, New York
 INTRODUCTION
  Upon inclusion of a hazardous waste site on the National Prior-
 ity List, various planning and design activities are initiated before
 remediation of the site commences. During the planning and design
 phases, relatively few people are involved and their health and safe-
 ty are protected through well defined measures specified in Health
 and Safety Plans. However, during the actual remedial work phase
 at a hazardous waste site, the health and safety concerns of an en-
 tire community  have to be addressed in a Contingency Plan. It is
 possible that emergency situations might arise during remedial con-
 struction which  would require the coordinated efforts of the con-
 tractor, local, state and Federal authorities to cope with the situ-
 ation.
  In this paper,  the  authors  discuss two contingency plans
 developed by the USEPA Region II for the Bridgeport Rental and
 Oil Services (BROS) site and the Lipari Landfill site, both sites
 located in Gloucester County,  New Jersey. In these plans,  the
 USEPA staff identified potential incidents which could impact the
 health and safety of the nearby community, who should be called
 upon to respond to these incidents, hazardous substances present at
 the site, specifications  of action levels in order to determine  the
 point at which the contingency  plan should  be activated and  the
 type of response action that should be taken.
BRIDGEPORT RENTAL AND OIL
SERVICES SITE

  The BROS site covers approximately 30 acres in Logan Town-
ship, Gloucester County, New Jersey. The site is a tank farm con-
sisting of 90 tanks and process vessels, drums, tank trucks and a
12.7 acre waste oil lagoon. The lagoon has an oily layer, aqueous
phase and bottom sludges.
  The USEPA began remedial measures  in the summer of 1983.
This  initial remedial work consisted of lowering the level of  the
lagoon by approximately 10 ft by removing 35 million gal. This
removal was accomplished by treating the aqueous phase of  the
lagoon and discharging it to Little Timber Creek. This action was
taken to prevent the lagoon from overflowing  its dike and to
stabilize the site until implementation of the long term  cleanup
program.
  During implementation of the initial remedial measure, a mobile
treatment system was constructed on-site. The treatment system
consisted of an oil/water separator, flocculation-sedimentation and
sand  and  granular  activated carbon  filtration.  Various lagoon
waste pumping systems were installed as well as fencing, decon-
tamination, personal hygiene and emergency medical facilities, etc.
  The contractor performing this work prepared a health and safe-
ty plan for his personnel. However, due  to unpredictable cir-
cumstances, a situation could have arisen where the air or surface
water in the vicinity of the site could have been contaminated. Such
a situation would have been beyond the control of the cleanup con-
tractor. The   incident  might  have  required  the   help  and
cooperatioan of local, county, state and Federal authorities to cor-
rect the situation.
  Therefore, a contingency plan was developed in conjunction with
all potential response participants. Part of the contingency plan
was devoted to preparing a telephone roster of the potential re-
sponding participants as well as area  hospitals. In addition, based
on the various physical* characteristics of the site, the chemical
characteristics of the lagoon (Table 1) and the type of work being
undertaken at the site, contingencies  were planned for the follow-
ing three emergency  situations:  Fire/Explosion,  Air  Quality
Deterioration and Water Quality Degradation.

Fire/Explosion

  A  fire on the site would have been handled by  the local fire
department with support given by other local fire departments and
area oil refinery fire fighting crews if necessary. Also supporting
this response would have been the contractor personnel who would
have taken appropriate measures to restrict the fire on site before
the fire department arrived. USEPA would have provided technical
assistance and additional air monitoring during this episode.
  During the  preparation of the Contingency Plan,  discussions
were held with the Fire Department and representatives of Logan
Township. During those discussions, it  was determined that the
Fire Department had neither the appropriate  equipment  to ade-
quately respond to this situation nor familiarity with the chemicals
at the site.
  In response to  their concerns, the  USEPA  provided  foam
generating equipment and self contained breathing  apparatus
(SCBAs) to supplement the local Fire Department's supply  (this
equipment is intended to be used at other Region II Superfund sites
after completion of  the current project). The Contingency Plan
also provided a list of the most significant chemicals and their con-
centrations (See Table 1), their hazard potentials,  harmful effects,
exposure limits, etc.

Air Quality Deterioration

  Due to the various chemicals known to be present at the site and
the unknown  nature of the material exposed as  the lagoon was
lowered, it was felt that a  situation could arise which could cause
the air  quality to deteriorate both on- and off-site.
                                                                                         SITE REMEDIATION
                                                       489

-------
                                                                                                        PROPERTY
                                                                                                        BOUNDARY

                                                                                                        MARSH
                                                                                                        TOPOGRAPHC
                                                                                                        CONTOUR
                                                                                                        TANKS AND
                                                                                                        VESSELS
                                                          Figure 1
                                                   General Site Arrangement
                                         Bridgeport Rental & Oil Services, Logan Twp., NJ
  As part of the health and safety plan, the contractor was required
to take air monitoring readings at intervals of every 4 hours using
an HNU photoionization detector. Periodically, Tenax or charcoal
tubes were utilized to measure air quality. If the air quality at the
perimeter of the site rose to 3 ppm above background for 15 min, as
measured by the HNU, the contractor would begin measuring the
air quality in the direction of the residents at locations 500, 3500
and 500 ft from  the site in succession.
  Once this off-site monitoring commenced,  Federal authorities
would advise the County Emergency Management Coordinator of
the situation and prepare to take action if needed. If the air  quality
at the off-site locations deteriorated to  3 ppm for 15 min above
background,  a recommendation would be made to the County
Emergency Management Coordinator to advise the residents of the
affected area to go into their homes and  shut windows, doors, etc.
  If the air  quality deteriorated  to 5  ppm  for  15  min above
background, a recommendations would be  made to the County
Coordinator to evacuate the homes. Names, addresses and phone
numbers of the citizens living hear the site were provided in the con-
tingency plan.
  However, any actual  evacuation decision would be made by the
local authorities. Also, local authorities would actually manage the
evacuation. The  USEPA would provide  technical assistance to the
local  officials  as well as extensive air  monitoring to assess  the
hazard and be in a position to recommened additional actions.
  The action  levels specified for  recommending  evacuation  of
homes were  based on the  USEPA's Interim Standard Operating
Safety Guide. The manual recommends that SCBAs be worn when
total  vapor  concentrations exceed 5 ppm  as measured  by in-
struments such as an HNU Photoionizer. Obviously, as more infor-
mation on the specific  type of contamination was gathered,  the
situation and previous decisions  could be reassessed. The use of 3
ppm above background as an action level to begin taking precau-
tionary  measures  was  primarily  based on conversations  with
members of various emergency response organizations.
                                                       Water Quality Degradation

                                                         As previously mentioned, the BROS lagoon's aqueous phase was
                                                       pumped,  treated and discharged into  a nearby  stream, Little
                                                       Timber Creek. Therefore, a potential existed for treatment process
                                                                                 Ttbte 1
                                                                Major Organic Chemicals Found in BROS Lagoon

Parameters Oil Phase
PCB
Ethylbenzene
Toluene
Aqueous Phase
Napthalene
Acenapthalene
Phenol
Benzene
1,1,1 Tr ichloroethane
Ethyl Benzene
Tetrachlo roe thane
Toluene
Tricholorethene
Bis-2-chloroethyl-ether
1,2 Dichloropropane
D i -e thy Iphtha late
Chlorobenzene

Ave
Concentrations
667 ppm
25 ppm
57 ppm

1 1 4 ppb
1 1 ppb
590 ppb
378 ppb
214 ppb
196 ppb
20 ppb
1394 ppb
87 ppb
29 ppb
1 1 0 ppb
493 ppb
11 ppb
85 ppb
490
SITE REMEDIATION

-------
failure and untreated discharge being released into the stream.
Also, a failure or rupture of one of the on-site tanks could cause a
release of contaminants to surface waters adjacent to the site.
  Upon detection of the discharge, the Corps of Engineers (COE)
resident engineer would notify the  USEPA, State and County
Health Department.
  These agencies would assume the responsibility for implementing
measures  to ameliorate, if possible, the contamination caused by
the spill and provide recommendations for further action. Also,
signs would be erected along the stream warning the public that the
stream had been contaminated and  that contact with the water
should be avoided.
LIP AM LANDFILL
  The Lipari Landfill site (Fig. 2), occupies approximately 16 acres
of a former gravel pit  in the Township of  Mantua, Gloucester
County,  New Jersey. Between 1958 and 1971, household wastes,
liquid and semi-solid chemical wastes  and other industrial waste
materials  were buried there.
  The site is bordered on  two sides  by two streams,  Chestnut
Branch and Rabbit Run, which converge near the landfill and enter
Alycon Lake approximately 1000  ft downstream of the site. Most
of the land area surrounding the  landfill is occupied by fruit  tree
orchards  and agricultural land. However, a housing development
of single  family occupied homes  is located approximately 500 ft
from the  northern boundary of the landfill. Some of the chemicals
found in  the leachate on the site are listed in  Table 2.
  As part of the first phase of cleanup at the site, a contract  was
awarded to construct a leachate containment system to encapsulate
the contaminated area of the site. The system consisted of a soil
bentonite groundwater cut-off wall, a synthetic membrane cap with
protective earth cover and a gas  venting system. Other ancillary
work included provisions for a support area, security and decon-
tamination operations, personal hygiene  facilities,  emergency
medical facilities, etc.
                            Table 2
       Organic Chemicals Found in Leachate of Lipari Landfill
  Parameters

  Bis  (2-chloroethyl)  ether
  Toluene
  Phenol
  Benzene
  Ethyl Benzene
  Methylene  Chloride
  Chlorobenzene
  Bis  (2-ethylexyl) phthalate
  1,2-Dichloethlyene
  Diethylpthalte
  1,1-Dichloroethlyene
  Trans-1/2-Dichloroethlyene
  Naphthlene
  1,1, Dichloroethane
  Trichloroetylene
  Acrolein
  Bis  (2-chloroisoproplether)
  Chloroethane
Concentration Range  (ppb)

   440-210,000
   51-22,400
   490-100,000
   9.8-2,012
   62-1,600
   .7-88
   11-106
   12-120
   43-5,800
   2.4-28
   4.4-36
   2.4-28
   5.7-102
   5.2-150
   0.9-23
   0-16
   0-80
   3-34
  Due to the chemicals present, existing conditions and the type of
work being performed at the site, it was felt that the only potential
hazards which the contingency plan should address were air quality
deterioration and fire.
Air Quality Deterioration

  Due to the trenching operation required for the construction of
the  slurry  wall,  there was  a possibility of releasing volatile
chemicals which could migrate off-site into a residential area. This
danger was believed to be especially critical when trenching  ap-
proached highly contaminated areas in the landfill which were
closest to homes.
  As part of the  contractor's health and safety plan, he was re-
quired to place  several Tenax tube air monitoring stations around
                                                           Figure 2
                                               General Location of Lipari Landfill
                                                                                              SITE REMEDIATION      491

-------
                           Figure 3
                Air Monitoring Station Locations
the periphery of the site. These air monitoring stations were design-
ed to collect samples daily, but the samples were not analyzed. In
addition,  monitoring stations were  located off-site  in the marsh
area between the site and the nearby residences.
  Also, the contractor monitored the air quality continuously at all
active work locations with an HNU photoionization detector. If
readings of 10 ppm or greater at the on-site active work area were
detected, the perimeter Tenax tube samples would be analyzed. In
addition, HNU detectors would be moved  to downwind perimeter
and off-site locations (Fig. 3).
                                                          The off-site sampling locations were situated near the residential
                                                        homes.  If total  organic vapor concentration readings at the site
                                                        perimeter locations exceeded limits similar to those set at the BROS
                                                        site, the County Emergency  Management Coordinator would be
                                                        advised by the USEPA or its representative of the situation and a
                                                        recommendation would be made to him to advise the residents to
                                                        remain  indoors, close windows,  etc.  The  County  Emergency
                                                        Management Coordinator was advised to keep a list of the names
                                                        and phone numbers of the potentially affected residents.
                                                          If these same total organic vapor concentrations limits were
                                                        detected at the off-site locations, then the Emergency Management
                                                        Coordinator would be advised to evacuate the affected residences.
                                                        Similar  to the BROS site,  after analyses of the Tenax tubes, more
                                                        specific information on the air contamination would be available
                                                        and the situation would be re-examined. The actual notification,
                                                        evacuation decision and evacuation would be performed by local
                                                        authorities. In addition, the USEPA and the State would provide
                                                        continuous air monitoring  as well as technical assistance to the local
                                                        authorities until  the situation was remedied.

                                                        Fire

                                                          Fire was deemed a  possibility at  this site.  Construction equip-
                                                        ment, fuel storage and the  possibility of striking a  flammable
                                                        substance during the trenching operation presented potential prob-
                                                        lems.  Similar  to the  BROS situation, the local Fire Department
                                                        would be notified and take command of the situation. Also, the ad-
                                                        ditional SCBA units and fire fighting foam stored at the BROS site
                                                        would be transported to the  Lipari Landfill  (approximately a 15
                                                        min driving distance) for the  Fire Department's use.  The USEPA
                                                        and State authorities would be made available to provide technical
                                                        assistance.

                                                        CONCLUSIONS

                                                          Contingency planning is  an essential activity that must take place
                                                        prior to the initiation of remedial action at a hazardous waste site.
                                                        Contingency plans must clearly delineate the potential emergencies
                                                        which may occur on- or off-site and identify various response par-
                                                        ticipants roles in these  emergencies.
                                                          One critical aspect in contingency planning is to make  certain
                                                        that all participants are educated on the potential emergencies and
                                                        hazards present  at the site. In addition, it is  critical  to assess the
                                                        response participants'  capabilities to  determine whether they are
                                                        prepared to  respond or whether additional  equipment, informa-
                                                        tion, etc. should be made  available to them.
                                                          By August,  1984, the majority of the work at the BROS site and
                                                        Lipari Landfill had been  completed. Fortunately, there has not
                                                        been a need to activate the contingency plans.  However,  careful
                                                        planning, including a  partial  testing of one of the plans, at these
                                                        sites would have provided  for a well coordinated response.
492
SITE REMEDIATION

-------
             SELECTING  SUPERFUND REMEDIAL ACTIONS

                                                D.  BRINT BIXLER
                                                J.  BILL HANSON
                                    U.S.  Environmental Protection Agency
                                                 Washington, D.C.
 INTRODUCTION

  The fourth year of the Superfund program has been marked by a
 shifting in the emphasis of USEPA activities at uncontrolled haz-
 ardous waste sites. During the initial years, the USEPA's primary
 goal was to identify the worst uncontrolled hazardous waste sites,
 investigate the nature and content of the problem at high priority
 sites and evaluate appropriate remedial actions.  As of July 31,
 1984, the USEPA has funded remedial investigations and feasibility
 studies (RI/FS) at 258 National  Priorities  List (NPL)  sites. As
 States and the USEPA Regions complete these RI/FS projects, the
 Superfund program is shifting from the initial study phase to the
 cleanup phase. The USEPA has already approved remedial action
 projects for the cleanup  of numerous sites. As of July 31,  1984,
 for the entire Superfund  remedial program,  which includes Fund-
 financed and private party response, there have been 134 cleanup
 actions approved at NPL  sites.
  The National Contingency Plan (NCP) defines three types of
 remedial actions.
 •Initial Remedial Measures (IRMs)
 •Source control measures
 •Off-site measures
  Currently,  the  USEPA's Regional  Administrators have  the
 authority to initiate all planning activities and to  select initial re-
 medial measures, but source control and off-site measures must be
 approved by the Assistant Administrator for Solid  Waste and
 Emergency Response at USEPA Headquarters. Before a remedial
 action is approved, the USEPA conducts a thorough technical,
 policy and legal review of the recommended project to ensure that
 NCP and CERCLA requirements have been met.


 PURPOSE

  A variety of technical information is required to select remedial
 actions. As USEPA's experience expands in this area, the informa-
 tion needs should become better defined. This paper identifies sev-
 eral of the key technical requirements needed before the State and
 USEPA can select the appropriate remedial action.
  The key to good decision-making is  the remedial investigation
 and feasibility study. The RI/FS is the single document that both
 identifies the nature and extent of contamination at the hazardous
 waste site and evaluates  feasible alternatives  to remedy the site
 problem. It is essential that the RI/FS contain information that is
adequate, both in quality and quantity, to allow  selection of the
correct remedial action. The quality and timeliness of the remedial
investigations and feasibility studies generally have not been ade-
quate. Too often, RI/FS projects take too long, present insuffic-
ient data, lack key alternatives and do not provide sufficient eval-
uation or rationale for the alternatives.
  Because of this problem, too many RI/FSs have to be modified
by  the State  or the USEPA, causing delay in selection of a re-
medial action. Some of the common problems encountered are:
•Likely alternatives are not identified early in the RI process
•Data are only minimally adequate
•The no-action alternative is not fully evaluated
•Incomplete cost estimates
•Alternatives  presented by outside groups are not evaluated
•On-site RCRA landfill requirements and groundwater correction
 requirements are not evaluated
  The USEPA has developed guidance documents on remedial in-
vestigations and feasibility studies that will avoid future problems
of this type. The Agency is also completing its policy on the appli-
cability of other environmental acts to CERCLA actions. Finally,
the USEPA is preparing a series of guidance documents relating to
specific types  of remedial actions. This series  will begin with drums
and tanks, surface impoundments and provision of potable water.
  This paper  discusses some key data and analytical requirements
for both remedial  investigations and feasibility studies needed be-
fore remedial  actions can be approved by the USEPA.
REMEDIAL INVESTIGATION

  The purpose of a remedial investigation is to define the nature
and extent of contamination at a site to the extent necessary to eval-
uate, select and design a cost-effective remedial action.  At most
sites, it would be prohibitively expensive to completely define the
nature and extent of contamination with 100% certainty. However,
the data collected must be sufficient to support an engineering eval-
uation of remedial alternatives. Unfortunately, many current RIs
are not focusing data collection properly;  as a result, costs are high
and studies lack specific detail to support the key remedies.
  The first step in an effective remedial investigation is to scope the
boundary of the project, identify the problems posed by the haz-
ardous substances and begin defining the type of remedial action
required.  If hazardous substances are still at or near the location
of original disposal, the RI/FS needs to focus  on  and evaluate
potential measures to  control the source of contamination (source
control measures). If the hazardous substances have migrated away
from the  original area, then measures must be developed to deal
with the contamination that has migrated in the environment (off-
site measures). In some cases, an RI/FS may address both types of
actions.
                                                                                          SITE REMEDIATION
                                                       493

-------
Key Data Needs

  Needed for source control measures is a definition of the hori-
zontal and vertical extent of on-site contamination. This definition
is necessary to determine how much control is required to prevent
or minimize migration of the contamination from the site.
  In some cases, the areas of contamination are easily defined.
For  example,   surface  impoundments usually  have  distinct
boundaries. The impoundment can be defined in terms of volume
and dimension and the waste characteristics identified. The im-
poundment must also be investigated to determine if liquids have
stratified; if so,  volume and characteristics  of each layer must be
evaluated. This  may  significantly  affect the development  of  re-
medial alternatives. For example, at one site, a layer of vinyl chlor-
ide was found at high enough levels to volatilize when exposed to
the atmosphere. This knowledge imposed a significant constraint
on the alternatives for dewatering and treatment of the wastes.
   Buried solid wastes or drums and tanks present  different source
control problems. The residual contamination  that has  migrated
into the soil adjacent to these more  concentrated wastes is more
difficult to define and evaluate. The boundaries of contamination
are usually not distinct; rather, the contamination varies with dis-
tance from the  source. A profile  of the concentration of these
contaminants must be determined by sampling soil at various dis-
tances from the source. Soil borings,  test pits or other appropriate
techniques should be utilized and  the samples  analyzed to allow
extrapolation of contamination profiles to background conditions.
Soil sampling may be stopped at the water table if additional ac-
tions are proposed to address contaminated groundwater;  however,
samples taken  in the saturated  zone are essential  in evaluating
alternative groundwater control or treatment methods.
   The analysis of soil contamination must show  the relationship of
vertical and horizontal distance  to concentration and volume of
contaminated material. This analysis is necessary to ensure that the
data are adequate to evaluate a  reasonable range of remedial  al-
ternatives.
   As with measures addressing the source of contamination, a pri-
mary function of a remedial investigation for contamination that
has migrated into the environment is to define the vertical and hor-
izontal extent  of contamination. The most  significant problem in
this area is contaminated groundwater which generally has a greater
area! extent but less  varied concentration than source contamina-
tion. Typically,  more extensive data collection  is needed to ade-
quately define the extent of off-site contamination. Whereas with
source control measures  the analysis of hydrogeologic conditions
is usually qualitative,  the  analysis must be quantitative when a
contaminated  groundwater plume is being  evaluated. The objec-
tive is not only to quantify the  source,  but also  to quantify the
movement of contamination in the environment. One or more key
contaminants must be selected as indicators, based on their per-
sistence and mobility in the environment and the degree of hazard.
   Aquifer parameters must be identified  so that  the rate and direc-
tion of flow can be determined. The data are needed both  for the
development and evaluation of remedial alternatives in the feasi-
bility study and  the subsequent engineering design. Effective data
collection during the remedial investigation will  minimize the need
for additional field work during the design phase.
   It is essential that the rate and direction of groundwater flow be
determined in  the investigation. In some cases,  this is a  relatively
simple procedure using the effective porosity, hydraulic gradient
and permeability of  the contaminated aquifer  and Darcy's equa-
tion to calculate groundwater velocity.
   At one site, the contaminated flow was confined by a subsurface
bedrock valley; therefore, the direction was known and the calcu-
lated  velocity  allowed a good  estimate of when contamination
would reach down-gradient wells. In  more complex settings, com-
puter modeling  is required to determine the movement of area-
wide groundwater plumes. One such site  has a series of four aqui-
fers beneath a concentrated source of contamination. Computer
modeling  was used to determine the rates and directions  of plume
                                                        movements  and to assess  the  impact of alternatives to control
                                                        migration.
                                                           The remedial investigation must also  identify the impacts of
                                                        plume movement on actual or potential receptors. This requires a
                                                        survey of the human and environmental receptors so that potential
                                                        for exposure to contamination can be identified. It also requires an
                                                        estimate of the contaminant level that may reach affected recep-
                                                        tors.  However, the numerous variables in the concentration, such
                                                        as dilution and attenuation of contaminants during migration, will
                                                        generally make a quantitative analysis too expensive and inaccurate
                                                        to determine the actual risk to receptors. Nevertheless, it is usually
                                                        possible  to make a qualitative estimate of exposure. In the case of
                                                        potential human  exposure, this  evaluation should focus on the
                                                        probability  that potential  receptors will  be exposed to contam-
                                                        inants at levels above existing USEPA standards, criteria or other
                                                        guidelines accepted as appropriate for the situation.

                                                        Accuracy of Data
                                                          The remedial investigation must include data of sufficient quality
                                                        and quantity to allow engineers to develop and evaluate alterna-
                                                        tives. Since cost is an important factor  in  selecting  a remedial
                                                        action, it is important that  the cost of all alternatives be estimated
                                                        to a level that will give the decision-maker  a realistic cost compari-
                                                        son. Moreover, the data must be of adequate quality to allow cost
                                                        estimates with a + 50Vt and  - 30% accuracy.
                                                          The engineer who designs the sampling program must consider
                                                        the data  needed to evaluate possible remedial alternatives and the
                                                        impact data collection has on cost estimation of those alternatives.
                                                        The accuracy of costs will depend on several factors, including:
                                                        •Accuracy of analytical data
                                                        •Estimate of waste quantities
                                                        •Estimate of waste strength and characteristics
                                                        •Accuracy of assigning costs
                                                          Each of these factors (and possibly others) contributes to varia-
                                                        tions in the costing of alternatives.
                                                          Each site should be reviewed to identify the factors having the
                                                        greatest  impact on  costs. This determination requires an assess-
                                                        ment  of  potential alternatives that may be evaluated in the feasi-
                                                        bility study to identify the data needed to  adequately cost  out the
                                                        alternative.
                                                          An adequate remedial investigation should identify all potential
                                                        alternatives early in the project and  determine the necessary data
                                                        needed to evaluate and cost those  alternatives in the feasibility
                                                        study. For example,  the quantity and type  of wastes in buried
                                                        drums can significantly affect treatment or disposal costs. There-
                                                        fore,  the sampling plan will include excavation of test  pits to sam-
                                                        ple a representative number .of drums. The quantity and condition
                                                        of the drums and the volume of  waste types should be estimated
                                                        so that an excavation  and off-site disposal alternative can be eval-
                                                        uated.

                                                        Summary of Investigation Data

                                                          A remedial investigation will typically generate a large amount of
                                                        data on various contaminants, locations and media polluted. This
                                                        information must be summarized to show the concentrations and
                                                        distribution of key contaminants for each medium investigated. An
                                                        assessment of the actual or potential for human and environmental
                                                        exposure will then be  made to determine if the uncontrolled waste
                                                        site represents a significant threat to the public or environment.
                                                          This exposure  assessment should include a  summary of the
                                                        significant contaminants for each medium and their character-
                                                        istics regarding movement in the environment, known or suspected
                                                        health effects, actual or potential pathways of migration and actual
                                                        or potential receptors. For groundwater contaminations, the ex-
                                                        posure potential should  be compared to  available USEPA stan-
                                                        dards, criteria or other accepted  guidelines. The exposure assess-
                                                        ment  is a key analysis that will be used to either approve or elim-
                                                        inate the "no-action" alternative in the feasibility study.
494
SITE REMEDIATION

-------
FEASIBILITY STUDY

  The NCP describes the overall process for selecting a cost-effec-
tive remedial action. There are four steps in the process:
•Development of remedial alternatives
•Screening out of infeasible alternatives
•A detailed evaluation of the remaining alternatives
•The selection of a remedial action
  The feasibility study converts the data on environmental con-
tamination collected in the remedial investigation into a series of
engineering solutions to the specific problem. Alternatives should
be developed for a range of technologies and cleanup objectives,
using acceptable standards, criteria and guidelines to  establish
baseline requirements. In the interest of cost and time, the feasi-
bility study should fully evaluate only a limited number of alterna-
tives. This section  discusses several  technical considerations and
requirements that have emerged from the USEPA's review of re-
cent RI/FS projects.

Development of Remedial Alternatives
   It is essential that the feasibility study develop alternatives over
a sufficient range of technologies and cleanup objectives to allow
the decision-maker to make an effective choice. Many completed
feasibility studies have not fully evaluated the appropriate alterna-
tives, thus causing delays in selecting a remedial action while addi-
tional alternatives were evaluated.
   As with remedial investigations, the feasibility study will usually
focus on alternatives that control the source of the migration of
contamination. For a typical source control measure, the following
alternatives should be evaluated:
•Excavation of wastes and off-site treatment and/or disposal
•Excavation of wastes and on-site treatment and/or disposal
•In-place containment
•No-action
   The excavation alternatives should usually consider several dis-
posal technologies. Off-site disposal must be at a facility that com-
plies with the requirements of RCRA and current USEPA policies.
When kpossible, alternative technologies that achieve the destruc-
tion of wastes (e.g., incineration) should also be considered. The
additional costs to destroy wastes,  rather than merely dispose of
wastes that would have the potential for future migration into the
environment, can often be justified by the increased reliability.
   On-site disposal alternatives must comply with the technical re-
quirements of the RCRA regulations at 40 CFR Part 264. This will
typically mean that a landfill in compliance with Part 264 Subpart
N should be fully evaluated.
   In-place containment alternatives  that prevent or minimize the
migration of contamination and the threat to public health and
the environment must be evaluated. This may  include a range of
technologies such as capping, slurry walls, fixation and stabiliza-
tion of wastes in place. Alternatives developed for on-site or in-
place containment should also consider technologies  that achieve
the destruction or treatment of wastes.
   Alternatives developed for contaminants that have  migrated
from the source must be capable of preventing or  mitigating ad-
verse public health  and environmental impacts resulting from
migration of the contamination in the environment.  Contaminated
groundwater is the major  off-site  problem. Typical alternatives
that will be evaluated include:
•Extraction  and treatment to reduce or remove contamination
•Barriers or gradient control to prevent further spread of contam-
  ination
•Provision of alternative water supply
•No-action
   The feasibility study should always evaluate a groundwater treat-
ment alternative that complies with RCRA regulations of Part 264
Subpart F. This regulation  requires extraction and treatment until
the groundwater concentration reaches either background levels or
maximum permissible  concentration limits. In  some situations,
alternatives  that contain the spread of contamination, rather than
removing it, must be evaluated. This step will be appropriate when
the contamination  has already spread over a  large area or the
source cannot be located or contained.
  Two key considerations will be: (1) the duration of ground-
water  control measures and (2)  the level  of  treatment required.
The duration must be estimated to allow the calculation of present
worth  cost. The level of treatment should be appropriate for the
intended use. Little or no treatment may be required if the water is
discharged to a wastewater treatment plant. However, if water will
be used for human consumption,  applicable public health stan-
dards or criteria should be used.

Screen Out Infeasible Alternatives
  During the screening step, alternatives may be eliminated from
further consideration for several reasons:
•Clearly inadequate protection of public health or the environment
 (for example,  the alternative does not prevent  migration of con-
 taminants)
•Comparatively high cost in relation to the protection and cost of
 other alternatives
•Significant adverse impacts of the remedy (for example, excava-
 tion of highly volatile, toxic wastes may have potential adverse im-
 pacts that cannot be mitigated)
  As part of the screening process,  requirements of other environ-
mental acts should  be assessed and the objectives of the remedial
action defined for each medium threatened.
  Remedial alternatives should prevent or minimize  both present
and future identified problems. Residual contaminants in the soil
or groundwater should be removed or contained  below standard
levels or at a concentration that will not pose a threat.
  Two to six alternatives generally  remain  after screening and are
examined in detail as part of the cost-effective analysis. These will
include the no-action and, for source contamination, construction
of an on-site landfill in compliance with RCRA.
Detailed Evaluation
  A detailed evaluation of the remaining alternatives is conducted
to evaluate their cost-effectiveness in addressing technical, environ-
mental and public  health concerns.  Environmental and public
health considerations can  often be readily evaluated in the context
of compliance  with appropriate  standards, criteria or guidance.
Hence, these factors are not discussed in detail here. It is the techni-
cal considerations which are often critical in selecting the alterna-
tive to be used.
                            Table 1
                 Typical Remedy Selection Factors
             Cost
             •Capital Cost
             •Annual Costs
             •Present Worth Cost
             Technical Considerations
             •Site Characterization
             •Reliability
             •Safety
             •Operation and Maintenance Requirements
             •Implementability
             Environmental Considerations
             •Adverse Effects
             •Beneficial Effects
             •Mitigation Measures
             Public Health Considerations
             •Populations Exposed
             •Route of Exposure
             •Type of Hazardous Substance
             •Level of Exposures
             Public Concerns
             •Public Acceptability
                                                                                                  SITE REMEDIATION
                                                            495

-------
                                                                   Table 2
                                                         Cost-Effectiveness Summary
Cost ($1,000)
Present Public Health
Alternative Capital Worth Considerations
1. No Action — — Unacceptable ex-
posure lo PAH if
summer or fire de-
mand requires use
of contaminated
well. Continued
water shortages.
2. Hookup to $250 $8.102 Reduces public
City X health threat 10 less
than 10-«.




3. Drill Deeper $1,870 $2,916 Reduces public
Wells health threat to less
than 10-6.







Environmental
Considerations
Continued migra-
tion of contam-
inated ground-
water; leading to
contamination of
Town Y'l water
supply.
Continued migra-
tion of contam-
inated ground-
water; leading to
contamination of
Town Y's water
supply.
Continued migra-
tion of contam-
inated groundwater
leading to contam-
ination of Town Y's
water supply. De-
pletes limited water
resource in deeper
aquifer.

Technical
Consideration.
...






Relies on simple
technology. No
treatment is
required.



Relies on proven
construction
technology.
Town A.






Publk
Comment Other
High resistance






Acceptable. Has significantly
higher OAM and
present worth con.




Acceptable to Town Has second highest
Z Park, but not to present worth
Town Y or cost.






         4. Aquifer
           Treatment
           A. Ozone
           B. Granular
           Activated
           Carbon
                        $374
                        $459
                        $709
                        $633
                          $1,618
                                     $2,109
                                     $2,434
                                     $2,150
At 2000 ng/1 of    Blocks migration
PAH, removes     and allows
taste and odor, but  additional wells to
results in 10 - ' lo   be opened.
10-6 risk.

At 1000 ng/l of
PAH, results in
10-5 to 10 -6 risk.
At 280 ng/1 of
PAH, results in
10-6 or less risk.

At 2000 ng/1 of     Blocks migration
PAH, removes taste and allows
Not used on wide   Acceptable.
scale. Less
rcsponsibe to slug
loading than GAC.
Would be expensive
to retrofit if treat-
ment goals change.
Certainty that
target levels will be
consistently met is
low due to opera-
tional inflexibility.
Present worth is
less than
GAC at higher risk
level but more at
lower recomment
goals.
                                      and odor but re-
                                      sults in 10-5 to
                                      10-6 risk.
                                                                  additional wells lo
                                                                  be opened.
                        $633
                        $633
                          $2,263
                                     $2,405*
At 1000 ng/1 of
PAH, results in
10-5 to  10-6 risk.

At 280 ng/1 of
PAH, result! in
10-6 or  less risk."
Considered best
available technol-
ogy. Dependable
over a wide range
of operating condi-
tions. Responds
well to slug loading
Likely to con-
sistently meet risk
target.
                                                                                        Acceptable.
Present worth is
less than other
technologies a
recommended
treatment level.
• Recommended Alternative
*• 280 ng/1 ii Ihe operational performance largcl for the GAC treatment system at this MIC The carcinogenic PAH will be reduced 10 a level less than or equal to 2.8 ng/1 u a remit of the operational performance
target. This will assure that the health riik to the population is lesa than or equal to a 10-6 health risk.
   Siting considerations, system reliability and effectiveness are key
technical factors.
   Site characterization: On-site containment may not be acceptable
if the hazardous wastes are in a poor location and the containment
structure cannot be designed to compensate for these shortcom-
ings of the site. Important locational criteria include:
 •Site stability
   •Flood prone areas: sites  within the 100  or  500 yr floodplain.
    Flood proofing could alleviate this threat; however, hazardous
    waste generally should not be contained in wetland areas.
   •Seismic zones: disposal facilities could be damaged by ground
    shaking  or associated  ground failure or subsidence; ground
    motion could result in differential settling and could also cause
    sediments to liquify, settle or slide.
                                                                •Landslide areas: a landslide can impact a disposal facility by
                                                                 carrying waste  materials downslope, by exposing waste ma-
                                                                 terials or by covering run-on/run-off controls.
                                                                •Subsidence-prone  areas:  over  a period of  years, long term
                                                                 faulting and surface deformation can occur as a result of sub-
                                                                 sidence; i.e., one Superfund site in Texas may eventually sink
                                                                 below sea level.
                                                                •High groundwater that moves rapidly and thereby may threaten
                                                                 current or future use of the groundwater.
                                                              •Ability to monitor and take future corrective action.  On-site con-
                                                               tainment may not be appropriate if the underlying  stratigraphy
                                                               is complex.  (Such as areas with  Karst geology or fractured bed-
                                                               rock.) A groundwater  plume  would be difficult or impossible to
                                                               monitor, and corrective action would be difficult to accomplish.
496
SITE REMEDIATION

-------
  Reliability: The reliability of a remedy can be evaluated in terms
of the complexity of operation and maintenance and the demon-
strated performance of the technology involved. Technologies re-
quiring labor intensive or complex activities are generally less re-
liable than systems with straightforward activities. But the most im-
portant aspect  of reliability is the ability of the remedy to meet
its long term performance goals. The method and result of failure
must be evaluated as part of the reliability analysis.
  Implementability: Remedies that can be constructed in a short
time and rely on demonstrated construction techniques are fav-
ored. Also,  remedies which do not threaten workers, the environ-
ment or the surrounding community during construction are fav-
ored.
 SELECTION OF A REMEDY
  All remedial actions undertaken under the auspices of CERCLA
 must be selected by the USEPA. All alternatives fully evaluated in
 the feasibility study will be presented to the decision-maker. A list
 of screened out alternatives will also be  presented. The feasibility
 study will not normally include a recommended alternative, since
 public comments on the alternatives  must be considered before
 selecting a remedy. Alternatives proposed by public or private par-
 ties that were not evaluated in the feasibility study will also be pre-
 sented.
  A number of factors must be evaluated before a remedial action
 can be approved. These include:
 •Cost
 •Technical considerations
 •Environmental considerations
 •Public health considerations
 •Public concerns
  Typical components for each category are shown in Table 1.
  The NCP requires that the USEPA select the "lowest cost al-
ternative that is technologically feasible and reliable and  which
effectively mitigates and minimizes damage to and provides ade-
quate protection of public health, welfare or the  environment."
Full technical compliance with all applicable and relevant USEPA
standards is the Agency's goal. However, limited exemptions may
be appropriate. For example, the USEPA might grant exemptions
on a site specific basis if the alternative that would be in technical
compliance  has excessive costs  or if site  conditions  make com-
pliance infeasible.  Therefore, alternatives  that may not comply
with USEPA requirements but which provide effective protection
of public health, welfare and the environment should be presented.
  In some  situations, use of USEPA advisories may result in a
range of alternatives. For  example,  treating contaminated water
supplies to achieve a range of carcinogenic risk levels may result in
similar alternatives with varying costs. These will be presented to
the decision maker.
  The USEPA has found a trade-off matrix useful to summarize
the alternatives and their relative advantages and  disadvantages.
A sample matrix is shown in Table 2. The information summarized
in this matrix must be fully explained in the feasibility study. In
some cases, all evaluation  considerations  will point  to  a  single
alternative. However, several alternatives will usually have compet-
ing advantages and disadvantages, making a decision more diffi-
cult.
  If the least cost alternative fully complies with USEPA require-
ments, it will generally be  selected. However, the USEPA also
prefers  alternatives  that  achieve treatment or destruction rather
than storage or disposal  of wastes. (This consideration applies to
both on-site actions and off-site facilities that will receive wastes re-
moved from a Superfund site.) In all cases, the USEPA must eval-
uate and document the selection process to ensure that its  decision
with respect to the various sites is generally consistent, but also re-
flects the specific conditions of each site.
                                                                                               SITE REMEDIATION
                                                          497

-------
           GUIDANCE FOR THE CONDUCT  OF REMEDIAL
                 INVESTIGATIONS/FEASIBILITY  STUDIES
                                    UNDER  SUPERFUND

                                       RICHARD L. STANFORD*
                                          JAMES LOUNSBURY
                              Office of Emergency and Remedial Response
                                 U.S. Environmental Protection Agency
                                             Washington, D.C.
                                           DOUGLAS AMMON
                             Solid and  Hazardous Waste  Research Division
                                 U.S. Environmental Protection Agency
                                              Cincinnati, Ohio
                                         S. ROBERT COCHRAN
                                           VIRGINIA HODGE
                                              JRB Associates
                                             McLean, Virginia
INTRODUCTION

  In this paper, the authors discuss the framework for the design
and implementation of an effective remedial investigation/feasibil-
ity study (RI/FS) project and focus on the most important aspects
of this framework. Although individual site conditions may vary,
the general framework developed here is flexible enough to accom-
modate these considerations and provide effective remedial plan-
ning at any  site. The framework described follows the National
Contingency Plan (NCR) (40 CFR Part 300).
  The remedial investigation provides an initial assessment of the
site conditions, identifies data needs,  defines the physical and
chemical conditions of the site that affect the implementation of
remedial measures and identifies current or potential threats posed
by the site. The RI emphasizes the collection and analysis of data
rather than the evaluation of the implications  of the data. Results
of RI  (Fig.  1) efforts produce a  data base  of source,  environ-
mental and impact characteristics  sufficient to evaluate the  need
for, appropriateness and/or effectiveness of  alternative remedial
actions.
  During the FS, the information gathered during the RI, and any
other relevant data, are analyzed  and evaluated to identify the
strengths and weaknesses of various remedial  alternatives. The FS
(Fig. 2) addresses five major characteristics of remedial alterna-
tives: (1) technical, (2) public health, (3) environmental, (4) insti-
tutional (compliance with legal requirements and policies) and (5)
cost.
  Even though the RI and the FS  emphasize  different aspects of
the overall response, the  two efforts are interdependent. The rela-
tionship of the RI to the FS in terms  of timing and findings is
shown in Figure 3. Interim reports that document data and  find-
ings generally are not formally required of the RI/FS effort. The RI
may be conducted in steps contributing to alternatives develop-
ment and analysis in the FS. These steps are discussed below.

INITIAL ACTIVITIES

  An important first step in an effective remedial investigation is
the review of existing data. The review should go beyond simply
describing the current situation; it  should also identify the scope
and procedures of subsequent investigations. These two important
efforts are shown as Tasks 1 and  2 of the remedial investigation
in Figures.

*Currcnlly wilh Clean Sites, Inc., Alexandria, VA.
                                                  Initial development of a background report detailing the nature
                                                and extent of the problem and outlining the purpose and need for
                                                further investigations and remedial actions is critical to effective
                                                remedial response. This initial assessment often  is used to develop
                                                response  objectives, determine the particular  requirements for
                                                further investigations and serve as the basis for enforcement activ-
                                                ities. This aspect will be discussed in more detail under Problem
                                                Definition and Project Scope.
                                                                                       Scoping and Planning
                                                                                       Processes Direct S
-------
                                                                            Screen Alternatives- Technologies
                                                                            NCR 40 CFR 1300 6eihllll.l2l.l3l
                                                                                                     Cost EHectiveneu
                                                                                                     Analysis NCP 40 CFB
                                                                                                     ! 30068(11(21
                                                             Figure 2
                                                      Feasibility Study Process
  Typical response objectives address contaminated aquifers and/
or water supplies,  remedy surface water contamination,  remove
direct contact hazard and other similar concerns. Once response
objectives have been determined, the scope of the remedial investi-
gation is determined. The work to be performed under the remedial
investigation should consider the quantity and quality of  existing
data, data needed to evaluate likely remedies and data needed to
determine present and potential health and environmental  threats.
This important use of existing data is often overlooked in develop-
ing remedial action plans.
  The initial assessment may also establish the basis for subsequent
enforcement activities. An important part of the initial assessment
is a review  of existing data on the levels of contamination in the
surrounding groundwaters, surface waters, air and drinking waters.
These data are compared to relevant public health and environ-
mental standards and criteria to establish a basis for requiring
private-party remedial actions. As exposure and  effects data are
evaluated during the RI/FS process, the initial report forms the
basis for a public health assessment.
  Often, initial assessments involve a site visit to confirm data re-
garding  site access, topography, exposed  populations and  other
information. The site visits and any data collection are performed
according to general safety and sampling plans. Once  the site visit
has been performed, detailed remedial investigations and manage-
ment plans are developed. These detailed,  site-specific plans can
be modifications of the more general plans. These plans, under the
site-specific work plan, include: the site safety plan, the quality
assurance/quality  control  (QA/QC) plan, the chain-of-custody
plan, the sampling plan, the  community  relations plan and the
site/data management plan.

PROBLEM DEFINITION AND PROJECT SCOPE

  Initial feasibility study activities build upon the initial data devel-
oped in the remedial investigation. The feasibility study efforts, at
this point, include a description of the proposed response and the
development of preliminary remedial technologies. These activities
are shown as Tasks 8 and 9 in Figure 3.
  The initial description  of the site characteristics from  the  RI
allows a preliminary identification of response actions  that could
be considered  as likely remedial  alternatives. These  general  re-
sponse actions do not necessarily identify specific technologies but
do identify generic activities  for  site remediation. Examples of
such generic activities include:
•Containment
•Removal
•Groundwater pumping and treatment
•Alternative water supplies
•Ground or surface water diversion
•In situ treatment
                                                                                                 SITE REMEDIATION
                                                                                                                              499

-------
  These general activities often comprise specific technologies that
can be readily evaluated for suitability at the specific site. When
this can  be done, specific  data that must be gathered during the
remedial investigation can be identified. If the specific technologies
cannot be identified at this stage, an effort should still be made to
identify specific data requirements.
  The description of  the  proposed response details general re-
sponse  objectives in terms of media affected,  populations con-
cerned, pathways of exposure to be remedied and likely technol-
ogies to be  employed.  A  clear statement of  response objectives
should be made to ensure  that a comprehensive response  is plan-
ned. Further, a detailed statement of response objectives  ensures
the identification  of all  data necessary  to evaluate alternative
responses. Thus,  the compilation of existing data, the identifica-
tion of data needs through a detailed statement of response objec-
tives and an initial evaluation of likely remedial technologies pro-
vide the information needed to determine the scope of an effective
RI/FS effort. An interim  report should be written at this point in
the RI/FS to provide interested parties a chance to review  progress.
SITE INVESTIGATION AND ALTERNATIVES ANALYSIS
  Site investigation activities in the RI are focused on characteriz-
ing the site with regard to public health and the environment, estab-
lishing baseline environmental and health conditions and  obtaining
data for use in the FS.  Simultaneously with the site investigation,
FS activities are initiated using data  made available  from the RI to
                                                                    analyze and to pinpoint applicable remedial alternatives through a
                                                                    series  of screening  and  analysis  processes.  These  activities are
                                                                    shown in Tasks 3 and 4 of the RI and Tasks 10, 11 and 12 of the FS
                                                                    in Figure 3. The objective of integrating the RI and FS activities at
                                                                    this stage of the site response is to efficiently use the resources com-
                                                                    mitted to site characterization efforts in the quantification of data
                                                                    to screen and evaluate remedial alternatives.
                                                                       Quantitative data obtained during the site investigation should
                                                                    include general source, pathway and receptor information used to
                                                                    screen proposed remedial alternatives. The data collected during
                                                                    this effort should include:
                                                                    •Environmental Setting. Data to define the site  and facility char-
                                                                     acteristics should be collected commensurate with the preliminary
                                                                     remedial  technology options of  interest.  Environmental  data
                                                                     should describe  the geography and layout  of  the  site and sur-
                                                                     rounding areas;  topography;  waste source locations, waste type;
                                                                     geotechnical  engineering considerations; normal   and unusual
                                                                     meteorological  conditions;  surface drainage patterns; geologic
                                                                     features;  groundwater occurrence,  flow  direction and rate; and
                                                                     soil type and chemistry.
                                                                    •Hazardous Substances.  Analytical data should be  collected to
                                                                     completely characterize the wastes including type; quantity; physi-
                                                                     cal form;  degree of contamination; disposition (containment or
                                                                     nature  of deposits); and facility characteristics affecting  release
                                                                     (e.g., site security, engineered barriers, etc.). These data may also
                                                                     be required to support decisions on interim remedial measures.
                             I
   IP           Work PUn
   Rapon       QA/OC Plan
Saa Background    Chan ol Conocty PUn
Maura ol ProbUm   Health it Safety PUn
tnanl of ProbUm   Management PUn
     of RaeponM  Sampeng PUn
               Communiry ReUuon* Plan
                                    R*m*dul Opliont
                                    N«goutlioru Document
                                                     Sow lot Banch b
                                                     P*ol Scab T.ili
                                                                                         f Ml RI Rapon
       i"* Raporu
Document Conuot
                                                 Public Hedllh
                                                 Assessment dit
                                                 Report
                             Hamadta! Invealiaauon

                             Mod* Scop* ol Wart
                                    lor
                             Remackal IrMtugiuani
                                                                                             0«li FS or RI *S Report
                                                                                          FeatUtUtty Study


                                                                                         Mo>W Stop* Ol Wort
                                                                                               lor
                                                                                          FcufeMy Slixta*
                                                                                                                      Rapon
                                                                                                                     Pau Oacure PUn
                                                                                                                     Compeance Uonuormg
                      Tut f I  Dwaipuan ol Current Silurian
                      Tut (2  PUni b Management
                                                                                 1***  n  De»o<)uon ol Pinpotart Reeponee

                                                                                 Tut  19  Pi«mmar> Rwmdul T*cnnulo0a*

                                                                                 Tut 110  OMlapm*ni ol AN«n*uvw

                                                                                 Tut 111  Imutl Sciwrvng at AJUTIUUVM
                      Tut /3  Sit* lnvMiig«iion

                      Tut 14  S*U InvwUQAuon AnAryt*


                      Tut ft  L*boiilary b Btncn Soul SludM

                      Tut M  fl*poru

                      Tut f!  Addmorvil R«quir«rmnu
                                                                                 Tut f 12 Andyt* al Alurntuvo
                                                                                 Tatt i 13 • Piehrrunery Rapon
                                                                                 Tut IW Final Report

                                                                                 Tut fib Additional Raqunmanu

                                                    • lui Ml I-S uiiUvt LluH CLHCLA 
-------
•Environmental Concentrations. Analytical data on air, soils, sur-
 face water and  groundwater  contamination in the vicinity of a
 site should be collected. These data should be sufficient to define
 the extent, origin,  direction and rate of movement of  contam-
 inant plumes. Sufficient background data should be collected to
 allow an assessment of hazards posed by the site in relation to the
 surrounding environment. Data should include time and location
 of sampling,  media sampled,  concentrations found, conditions
 during sampling and  the identities of individuals performing the
 sampling and analysis.
•Potential Impact on Receptors. Data describing the human pop-
 ulations and environmental systems that are  susceptible to con-
 taminant exposure via the transport pathways from a site should
 be collected to  assess present or potential exposures. Chemical
 analysis of biological  samples may be needed. Data on observable
 effects in ecosystems may also be obtained.
•Remedial Action Effectiveness. Data relevant to  the feasibility
 and effectiveness of proposed remedial actions should be collected.
 Because of the diversity of potential alternatives, specific investi-
 gations of this nature may be delayed until the conclusion of rele-
 vant portions of the feasibility study.
  As a result of the site characterization activities, several types of
assessments can be performed. Three types of  typical assessments
are: (1) a contamination assessment determining how the  contam-
inants are distributed  and moving; (2) a public health assessment
determining the human health implications and  establishes baseline
conditions for potential litigative action; and (3) an environmental
assessment determining environmental implications  of the site. In
achieving the goals of the site investigation and associated assess-
ments, several types of technical investigations are required. These
investigations may include source characteristics, geologic  and
pedological conditions, groundwater characteristics, surface water
transport mechanisms, atmospheric conditions, identification of
contaminants of  concern, human health effects and biological and
ecological effects.
  A thorough analysis and summary of all site investigations and
their results should  be prepared to  ensure that the investigation
data are sufficient in quality and quantity to support the feasibility
study. The  data  from all site investigations should be organized
and presented  logically to clearly show  the relationships  between
site investigations for each medium.
  During and immediately following the site investigation  effort, a
feasibility study is directed at developing a limited number of alter-
natives for source control or  off-site remedial actions. These alter-
natives are developed in response to the  objectives of  the site
remediation efforts and the results of the RI.
  A critical step  is the establishment of site-specific objectives for
the response. These objectives  will be based on public health and
environmental  concerns, the description of the current situation
(from Task 1), information gathered during the remedial investiga-
tion, Section 300.68 of the  National Contingency Plan (NCP),
applicable USEPA RI/FS guidance  and the requirements of any
other applicable  USEPA,  Federal and State environmental stan-
dards,  guidance  and  advisories as defined under the USEPA's
CERCLA compliance  policy. Objectives for source control meas-
ures should be  developed to  prevent or significantly minimize
migration of contamination  from the site. Objectives for off-site
measures should  prevent  or  minimize impacts  of contamination
that has migrated from the  site. Preliminary  cleanup objectives
should be developed in consultation with the USEPA and the State.
  Once response objectives are established, a  series of remedial
alternatives are identified. These alternatives incorporate remedial
technologies (from Task 9), response objectives and  other appro-
priate considerations into a comprehensive, site-specific approach.
There may be overlap among the alternatives developed, and addi-
tional data may be required to fully develop the alternatives. Alter-
natives should be developed in close consultation with the  USEPA
and the State; the rationale for excluding technologies earlier con-
sidered  applicable should  be  documented. During  this part of the
remedial response, the RI and  FS activities need to be well coor-
dinated; information should  be flowing freely from one to the
other.
  After a series of remedial alternatives has been developed, it is
often necessary to reduce the remedial action options to a manage-
able number. Initial screening of alternatives is performed using
three broad considerations as  performance  factors: public health,
environment and costs. In detail these factors include:
•Public Health and Environmental Protection. Only those alter-
 natives satisfying the response objectives  and contributing sub-
 stantially to the  protection of public  health, welfare  or the en-
 vironment should be considered further. Source control alterna-
 tives should achieve adequate control of source materials. Off-
 site alternatives should minimize or mitigate the threat of harm to
 public health, welfare or the environment.
•Environmental Effects.  Alternatives posing significant adverse
 environmental effects should  be excluded.
•Cost. An alternative whose cost far exceeds that of other alter-
 natives will usually be eliminated unless significant benefits will
 also be realized. Total  costs will include the costs for implement-
 ing, operating and maintaining the alternatives.
  At, or near the conclusion of, the RI site  characterization activ-
ities, there should  be sufficient data to complete a detailed evalua-
tion of these alternatives. Alternative evaluation efforts in the FS
should include:
•Detailed development of the alternatives
•Environmental analysis
•Public Health analysis
•Institutional analysis
•Cost analysis
  Alternatives should be evaluated using  technical environmental
and economic criteria. At a minimum, the following areas should
be used to evaluate alternatives:
•Health information
•Environmental effects
•Technical aspects of the remedial alternatives
•Present worth of total costs
•Information on technical and environmental standards and ad-
 visories
•Information on community effects
•Other factors
  The results of this evaluation and any material documenting fac-
tors directing the evaluation efforts are passed on for incorporation
in the preliminary report.

REFINEMENT OF TECHNOLOGY EVALUATION
  Bench and pilot scale studies may be necessary to obtain suffic-
ient data for the selection and implementation of remedial action
alternatives. As shown in Figure 3, bench and pilot scale studies are
part of the RI task sequence (i.e., Task 5) and generally should be
conducted  concurrently with the later stages of the analysis of al-
ternatives in the FS (i.e., Task 12). The interaction between these
two tasks is important. The analysis of alternatives may require the
bench and pilot scale  information to determine the feasibility of a
given technology since there may be a lack of  long-term perfor-
mance information about remedial action technologies and site and
waste heterogeneities.
  Bench scale studies differ from pilot scale studies in purpose,
size and application. Bench scale studies are  much smaller in scale,
cost, time and waste volume. Their purpose is to determine appli-
cation feasibility over the expected  range  of conditions. Bench
scale studies are flexible; a wide range of variables can be eval-
uated when determining the performance limitations and capabil-
ities of a technology.
  Pilot scale studies, on the other hand, are used to guide the selec-
tion of one alternative  from  several (when this cannot be done
appropriately at the bench scale) and to define the design, oper-
ating criteria  and  specific features of an alternative likely to be
selected. Pilot scale studies can be used to determine the stability of
a process or material (e.g., compatibility tests) in an application
                                                                                                SITE REMEDIATION
                                                           501

-------
                                                                  Table 1
                                              Examples of Bench and Pilot Scale Testing Programs
                                 d la I  Technology
                                                                                    r«a«pie
                              Air Po 11 lit Ion and Ga* Hiqr * t ion  Con
                              1.  Capping
                              2.  Duct Control
                              3 .  Vapor Col Lection and Trea uient
                                 (carbon ad*orptlont


                              Surface Water Control!
                              1.  Capping
                              2.  Grading
                              3.  Reveqe ta t_ion
                           C.  Leachate and Ground*at*r Control*
                              i.  Containment barrier* <*lurry «* 11 • .
                              i.  Groundvater puapmq (w*ll point*,
                                 •uctlon »*mll* etc. )
                              ).  Subturface collection drain*
                              4.  Pemeable treatment bad*  (Umetone, AC)
                              5.  Capping
                           D. Dlr*ct w««t« Control
                             i .  Incineration
                             2.  SolldlflccUon
                             ).  BloloqlctL
                                 fcctl v* ted
                                 Trictcllnq Fllt*n
                                 Ch«Blc«l
                                 Omd«t ion/Reduction
                                 Prvcipi titlon
                                 Ion exchange r««ini
                             *>.  Phyiictl
                                 C*rbon Adiorption
                                 PlocculAtlon,  Pr*cLplt*tion, S*di««nutlon
      OL**olv*d Air notation
      Mr  Stripping
      wet  Air ORIdatIon
   ft.  In Situ Treatment
      Hlcroblal Degradation
      Neutralltatlon/DetOKifIcation
      Precipitatlon
      Nitrification
   T.  Land 01*po*al [landfill,  land application)

t.  Soil and Sediment Containment And Removal
      Excavation
                                Capping
                                                             ft4ncn,  Soil density  *nd bearing •-•p*ctty v*.

                                                               ** r « rlal«.
                                                             Pll
                                                                      -Pi
                                                               »«w i  411-iriav.w .u. » vj-... . - . --	_..._,.-,-.
                                                               of ga* withdrawal  rate* to  control  re leaiee.


                                                             Bench:  Horticultural ceat Ing to **»Lect grave
                                                               • pmcla*  end toi i. conditioning for vegetative
                                                               cover.
                                                             Pilot;  In-pl*ce teitlnq of qeoteitilea for
                                                               control  of eroalon in graaaed diversion
                                                               11tehee.
                                                                       Dan
                                                                         vi  grain «lz« of 11 me atone material  for a
                                                                         treatment b*d.  s«t« r»ln*tlon of -^*«ic«l
                                                                         compatibility o( « co«p*ct*J  cheeiica I
                                                              com position
                                                            Pilott  Conduct qeopfiy«ical *^rv«y to *ap
                                                              contalnatent pi ^e depth and eitent.  Con-
                                                              duct uqnetOMter iurv«y to located buried
                                                              metallic object*.
and are aimed at delineation of specific design and operating cri-
teria.
   The  diversity of activities that may be required in the selection
and application of a remedial technology is shown in Table 1. The
examples of bench and pilot scale studies presented illustrate the
diversity of disciplines  and sciences required to define application
conditions for the technologies.


DRAFT RI/FS REPORTS

   The final RI report is shown as Task 6 in Figure 3. This report
should  include the results of Tasks  1 through 5 and should include
additional information  in appendices. A recommended format has
been established; it is structured: (1) to enable the reader to easily
cross-reference data; (2) to ensure that all  major issues are ade-
quately  addressed;  (3) to promote high quality and consistency in
RI studies; and (4) most  importantly, to ensure adequate  docu-
mentation and completeness of  data  entering into the decision-
making process.
   The Public Health Assessment  Report is prepared as needed for
enforcement cases. This report is prepared earlier in the RI study
as an evaluation of exposure and  risk to human health and the en-
vironment,  drawing on  data and analysis from Task 4, the Site In-
vestigation Analysis, as  shown in Figure 3.
                                                                The draft final FS Report (or combined RI/FS Report) is shown
                                                             as Task 13 in Figure 3. In this report, the engineers should summar-
                                                             ize the results of Tasks 8 through 13 and should include any sup-
                                                             plemental  information  in  appendices.  Again,  a  recommended
                                                             format has been established to permit the evaluation of feasibility
                                                             studies on a common basis and shorten the lead time required for
                                                             the  selection and implementation of  the cost-effective remedial
                                                             action strategy.
                                                                A draft FS report provides a decisionmaker with important in-
                                                             formation for choosing the  remedy  providing the best balance be-
                                                             tween cost and health protection, environmental protection and en-
                                                             gineering reliability. A minimum of 21 days is also provided after
                                                             completion of the feasibility study for the public to comment on the
                                                             alternative  analysis. Based  on the  public comment,  revisions to
                                                             some or all alternatives may  be appropriate.  A response of this
                                                             summary is prepared in the final revision of the FS report to docu-
                                                             ment public comments and to describe the actions taken  regard-
                                                             ing these comments.

                                                             ADDITIONAL REQUIREMENTS

                                                                Tasks 7 and 8 in  Figure 3 indicate that additional requirements
                                                             may  be needed. These  include  items such as administrative  re-
                                                             ports, document control procedures, community relations support,
502
SITE REMEDIATION

-------
post-closure plans,  compliance monitoring schedules and other
situation-specific requirements.
CONCLUSIONS

  An organized management approach to remedial investigations
and feasibility studies  should avoid  many of the problems en-
countered in less well-planned studies and thus lead to successful
cleanup. This paper has summarized detailed guidance that will be
forthcoming from the USEPA.
  The information in this paper is based on August,  1984, draft
guidance for conducting  remedial investigations  and feasibility
studies under CERCLA. The information is subject to change and
does not necessarily reflect official USEPA guidance. The reader is
advised to consult with the USEPA for revisions to the guidance.
ACKNOWLEDGMENT

  The Remedial Investigations and Feasibility  Studies guidance
development projects were supported by the following contractors:
Anderson Nichols & Co.
Battelle Pacific Northwest Laboratories
Battelle Project Management Division
Booz-Allen & Hamilton
CH2M Hill
Engineering-Science
Environ
Environmental Law Institute
ICF Inc.
JRB Associates
NUS Corporation
Radian Corporation
Versar
  The  dedication of the many participants from these contractors
is acknowledged.
  The  authors would like to acknowledge  the efforts of many
USEPA work group participants and reviewers who contributed to
this project. The efforts of Brint Bixler,  Bruce Clemens, Sylvia
Lowrance, Roy Murphy, Larry Raniere, Jim Spatterella, and Craig
Zamuda are especially appreciated.
                                                                                             SITE REMEDIATION
                                                                                                                         503

-------
       CLEANUP  OF  RADIOACTIVE MILL  TAILINGS  FROM
                     PROPERTIES  IN MONTICELLO,  UTAH

                                             FLOYD D. NICHOLS
                                                JOHN M. BRINK
                                             PHILIP C.  NYBERG
                            U.S. Environmental Protection Agency, Region VIII
                                                Denver,  Colorado
INTRODUCTION

  In conjunction with an early 1970s USEPA investigation of pos-
sible tailings misuse at a former U.S. Atomic Energy Commission
(AEC) uranium millsite in Monticello, Utah, two sites were tenta-
tively identified as being radioactively contaminated. The sites were
an occupied residence  on a 0.6-acre lot and a small downtown
store, both built during the mid-1940s.
  Studies carried out by the USEPA,  the U.S. Department of
Energy (USDOE) and  the Utah State Health Department in the
early 1980s revealed radon progeny and gamma radiation levels in
the structures that were greatly in excess of generally accepted
health criteria. Believing the radioactivity had originated from the
AEC operation,  USDOE carried out engineering assessments for
both buildings can concluded that the source of the contamination
was radioactive mill tailings incorporated into the  adobe walls of
the buildings and used as fill underneath or adjacent to them. How-
ever, a search of the build records revealed  that the mill tailings
had evidently been obtained from the vanadium processing facility
which predated  the AEC  uranium plant.  This meant  that the
USDOE, successor agency to the AEC, did not  have authority to
conduct any remedial activity, and this effectively suspended pro-
gress at the sites until USEPA action under the Superfund program
was initiated in  1981. A "Superfund"  cleanup was completed
under authority of CERCLA during the summer of 1984, and the
restored properties were reoccupied by their current  owners.

History

  The historical  beginnings of uranium and vanadium processing
in Monticello date back to the latter part of the 19th century, when
carnotite, an  ore containing uranium, radium and  vanadium, was
discovered in the Colorado Plateau area of southeastern Utah and
southwestern Colorado. The extraction and processing of these ma-
terials followed cycles that were determined by  markets for them
and their availability from alternate sources. For  example, explora-
tion for and  processing  of carnotite reached a peak during the
early 20th century when the Denver Radium  Institute was pro-
ducing radium for research and medicinal purposes. This market
subsequently  collapsed with the discovery of extensive, rich de-
posits of ore in what is now Zaire.6
  Another peak  in carnotite mining began in the late 1930s when
the outbreak  of war in Europe  stimulated interest  in assuring the
availability of adequate domestic sources of canadium and other
minerals. Vanadium is a non-radioactive metal used as a hardener
in steel, and  the  federal  government—through  its depression-era
agencies and authorities—played a major role in creating and con
trolling the markets for it.
                                                     The Monticello mill processed vanadium as a part of the World
                                                   War II production effort until 1944 when sufficient stockpiles were
                                                   produced. The closure of the mill was only temporary, however.
                                                   It was reopened in 1946 and used to produce uranium for the U.S.
                                                   Government's World War II atomic weapons program (the Man-
                                                   hattan Project) and post-war Atomic Energy Commission projects.
                                                   The mill continued to produce uranium for the defense effort until
                                                   1960, when it was  closed and the  buildings were dismantled.
                                                   USDOE still controls the millsite under its Surplus Facilities Man-
                                                   agement Program. The one million tons of tailings in several stabil-
                                                   ized piles are all that remain there. Although this mill is now his-
                                                   tory, mining and milling of uranium for nuclear power applications
                                                   continues in southeastern Utah to this day.
                                                     At the time the two buildings  were constructed in 1944-45, there
                                                   was little or no appreciation for  the potential health hazards of the
                                                   radioactivity in the vanadium mill tailings. There were no controls
                                                   on the use of  these  waste materials at that time except how they
                                                   might best be stored at a minimum cost. In the 1960s and '70s, the
                                                   USEPA and its predecessor agencies conducted field surveys that
                                                   identified the  presence  and magnitude  of the uranium mill tail-
                                                   ings dispersal  problem,  primarily in the  western United States.
                                                   These studies played a major role in the passage of the Uranium
                                                   Mill Tailings Radiation Control Act  of 1978 (P.L. 95-604). This
                                                   Act empowered  the Nuclear Regulatory Commission to regulate
                                                   present and future uranium milling operations according to stan-
                                                   dards set by the USEPA to prevent this  problem from reoccurring
                                                   in the future. In addition, it authorized the USDOE to clean up cer-
                                                   tain designated inactive uranium millsites to meet USEPA stan-
                                                   dards. With this Act, the Government recognized its obligation to
                                                   remedy problems created by the production of uranium for the de-
                                                   fense effort, but left unresolved the problems created by the pro-
                                                   duction  of similar wastes generated prior to the "Atomic Age."
                                                   Where  such problems exist—as in Monticello—the use of Super-
                                                   fund may be an alternative for resolution.
                                                   HEALTH IMPLICATIONS OF RADIATION EXPOSURE
                                                     The mill tailings used in the construction of the two buildings in
                                                   Monticello contained up to 280 pCi/g* radium (Ra-226). Radium
                                                   is one of a number of radioactive elements which comprise the
                                                   daughter products of uranium (U-238). In fact, 13 different radio-
                                                   active elements are formed in sequence as the uranium ultimately
                                                   decays down to stable lead, as shown in Figure 1. Radioactive ele-
                                                   ments, or radionuclides, emit gamma rays and decay by the emis-
                                                   sion  of alpha particles  and beta particles. These three—alpha,
                                                   beta  and gamma radiations—are referred to as ionizing radiations
 504
SITE REMEDIATION

-------
in order to distinguish them from less energetic particles or electro-
magnetic radiations such as microwaves and light or radio waves.
  Alpha particles are physically identical to helium nuclei, contain-
ing two protons and two neutrons; they are the most ionizing but
least penetrating of the common radiations. Outside the human
body, they represent a negligible health threat as they cannot even
penetrate the epidermal layer of the skin.  If inhaled or ingested,
however, alpha particles may become a significant human exposure
problem.
  Beta particles are, in reality,  high-speed electrons originating in
the nucleus of a decaying, unstable atom. They  are less ionizing
than alpha particles,  but are more penetrating.  The more ener-
getic beta particles found in nature can penetrate up to a few centi-
meters of tissue and a few meters in air.
  Gamma rays are not particles at  all,  but are  highly energetic
photons, similar to x-rays, which can easily penetrate several centi-
meters of tissue and are considered  a hazard for radiation ex-
posures both external as well as internal to the body. All three of
these radiations are produced in the decay of either radium or its
daughter products.
  Of all the radionuclides in the uranium decay series, radium and
its immediate daughters are the  ones of greatest human health con-
cern. Radium, with  a half-life of 1,620 years, will  remain at its
present concentration in the environment  for  long periods com-
pared to a human  lifetime unless it is physically removed. Radium
by itself, however, is not of such great concern as its  decay pro-
ducts, viz., radon (Rn-222), polonium-218 (also called radium-A),
lead-214 (called radium-B), bismuth-214 (called radium-C)  and
polonium-214. Radon is  unique among this group because it is a
noble gas that is inherently mobile  at normal temperatures  and
pressures, while all the others  are solids. This gaseous nature of
radon is what makes a radium contamination problem  so trouble-
some, for there is, thus, an inevitable, built-in, airborne dispersal
mechanism contained in the radium decay chain.
   Furthermore, the decay products of the radon have very short
half-lives, making them very radioactive. When radium decays to
radon, which has a half-life of about 4 days, the resulting gas has a
good opportunity to diffuse out of the material in which the radium
was contained and thence into  the atmosphere before undergoing
subsequent decays.
   If the radon is contained within a volume and not  allowed to
disperse, the airborne concentration of the radon daughters can
build up to levels that may represent  a significant health threat if
inhaled over an extended period of time. This is, in fact, what hap-
pened in some of the uranium mines before the problem was recog-
nized and mine ventilation was required by law.
   The health hazard posed in  the short term  by radiation from
radium and its daughter products in concentrations usually found
in the environment is normally  not life-threatening, as the radium
is almost never sufficiently concentrated to cause acute radiation
syndrome or other prompt radiation effects. Rather, the hazard is
generally considered to be a greater likelihood of cancer induction
at some future time,  with the probability of cancer directly pro-
portional to radiation dose.
  The human exposure from radium-bearing materials  is twofold:
first, the external exposure to gamma radiation from some of the
radium decay products; and second, the internal exposure to the
lungs because of the inhalation of the short-lived radon daughter
products which emit alpha particles. Of these two, the second is by
far the most significant, representing the largest radiation exposure
normally received by the general public from any  source.3  The
health effect of greatest concern, therefore, is lung cancer induc-
tion, and it has been estimated that between 2000 and 20,000 of the
120,000 annual lung cancer deaths in this country may be the re-
sult of exposure to radon daughter products.'12> 4
  Standards for the regulation of gamma radiation exposure to
radiation workers have been set by the Nuclear Regulatory Com-
mission; allowable limits to the general public were issued by the
•  pCi/g = picocuries of radium per gram of soil; a picocurie is a fractional portion of a curie, a
       unit of radioactivity equal to 3X10'° disintegrations per second.
Federal Radiation Council, whose responsibilities subsequently
transferred to the USEPA. The standards are expressed in terms of
the maximum allowable dose per year in units of "rems"*. Radia-
tion workers are allowed to receive up to 5 rems/yr (10 CFR 20),
while individuals in the general  public may receive up to !/2  rem/
yr',  exclusive of  natural background  radiation or any medical
radiation exposure. For comparison purposes, natural background
radiation in the Monticello, Utah area  is estimated to be about
0.14 rem/yr. Based on measurements made in the two buildings
addressed here,  people living in the residential structure could re-
ceive up to 0.33 rem/yr based on an assumed occupancy 75% of the
time. People working in the store could have received up to  0.2
rem/yr based upon an assumed average work time of 40 hr/week.
  The standards for controlling exposure to the short-lived radon
daughters are not as well defined as those for gamma radiation ex-
                            Figure 1
                   Uranium-Radium Decay Series


posure. The units are different as well, often being expressed as
"working level-months"  (WLM), where the working levelf is a
measure of the radon daughter concentration expressed as the po-
tential alpha particle energy contained therein. A WLM is defined
as the exposure an individual would receive in an atmosphere con-
taining one (1) WL if he remained there for 170 hr,  an average
working month. The Mine  Safety and  Health Administration of
the U.S. Department of Labor set a  maximum  limit of 4 WLM/
year for persons working in uranium or other mines but did not
address exposure limits for  the general public.  The USEPA  has,

*     rem: The unit of dose equivalent, which is equal  to the product of the absorbed radia-
        tion dose, the quality factor for different  LET radiations, and any other modifying
        factors.

t Working level: a special unit of exposure to short-lived radon decay products in air; one Working
Level equals 1.3 x 10'MeVof potential alpha energy from any combination of radon daughters per
liter of air.
                                                                                                 SITE REMEDIATION
                                                            505

-------
L
                           Figure 2
  Floor Plan Showing Installation of Electrostatic Precipitaiors (ESPs) in
                            Store
                 11/11

                 MOM.
                            Figure 3
       Radon Daughter Reduction Due to Electrostatic Precipitator
however, set radon daughter exposure standards for certain specific
cases which are quite similar to the Monticello situation.
  Under the Uranium Mill Tailings Radiation Control Act, the
USEPA set standards for the cleanup of certain designated in-
active  uranium mill tailings sites  and any nearby properties con-
taminated  with tailings (40 CFR  192). According to these stand-
ards, the  maximum radon daughter  concentration   (RDC) per-
mitted in a habitable building contaminated  with  uranium mill
tailings is 0.02 WL based on an annual average determination. For
comparison, the authors estimate that the typical U.S. home has an
average of 0.005 WL.
  Before rehabilitation, the residential structure in Monticello had
an  average concentration estimated at  0.1  WL, five times the
USEPA standard. The store was even higher with  an estimated
average concentration ranging between 0.2 and 0.3 WL; ten  to
fifteen times  the USEPA standard. While the USEPA standards
do not legally apply here, these standards were adopted as the min-
imum  goals for the  cleanup project because the situation was
analogous to that contemplated in 40 CFR 192.
  The  standards found in 40 CFR 192 are useful in deciding when
the gamma radiation exposure rate or the indoor radon daughter
concentration is too high. The actual health risk from this exposure
represents an important consideration in the standards setting pro-
cess. The adverse health  effect expected from  excessive radiation
exposure is an increase in the probability of cancer in the future.
There are many studies involving both human and animal subjects
which  have been  used to define the relationship between health
effects and radiation  dose. Most health  effects, however,  were
                                                        noted at much higher doses than found in this current situation, so
                                                        extrapolation from higher to lower doses is necessary in order to
                                                        predict the probability of future health effects. There is, of course,
                                                        some uncertainty introduced by this extrapolation,  but  it must be
                                                        recognized that what is being discussed is the statistical probability
                                                        that an event will occur. Any such prediction has some uncertainty
                                                        inherently  associated with  it when applied to specific  individual
                                                        radiation doses.
                                                          As mentioned  earlier, the dominant  risk posed by radioactive
                                                        tailings is an increase in the probability of lung cancer from  the
                                                        inhalation of radon daughter products. According to USEPA stud-
                                                        ies, 3<4 a lifetime of exposure in a residence (assuming 75% occu-
                                                        pancy) with an annual average RDC of 0.1 WL represents an in-
                                                        creased risk of lung cancer of about 9%.
                                                          The chance of incurring  a fatal  lung cancer from any source
                                                        (other than radiation) is about 3% in the U.S. population. This
                                                        effectively means the average person typically has 3 chances in 100
                                                        of incurring a fatal lung cancer.
                                                          The added risk incurred by a lifetime of residence in the Monti-
                                                        cello, Utah home was about four times the normal risk.  Although
                                                        the  store had a higher annual average RDC than the house, the in-
                                                        creased risk of a lifetime of working in the store is about the same
                                                        as living in the house because of the fewer hours per day normally
                                                        spent there. While this may not seem like a significant risk at first
                                                        glance, it must be realized that a lifetime of living or working in
                                                        these two buildings effectively quadrupled the  occupants' chances
                                                        of incurring lung cancer.
                                                                                                        '•••MO 101        I
                                                                                   Figure 4
                                                                   Plan View of Site, Showing Contaminated Areas
                                                                          o
                                                                  0
0

                                                                                   Figure 5
                                                          Plan View of Residence and Lot, Showing Location of Contamination
 506
SITE REMEDIATION

-------
RESPONSE AND REHABILITATION

  In early 1981, the USDOE contracted for an engineering assess-
ment and remedial action plan for the two most contaminated
structures in Monticello; a mail-order catalog store and a private
residence. The resulting assessment provided detailed plans for
demolition and reconstruction of the two structures and  removal
of the contaminants from the sites. Before work could begin, how-
ever, the USDOE determined that it lacked the legal authority to
clean up the properties and suspended any further mitigation ac-
tions.
  After it was determined that the USDOE did not have authority
to clean up the two sites, the State of Utah and USDOE requested
that the USEPA clean them up under CERCLA. CERCLA pro-
vides three main avenues of response to releases or potential re-
leases of hazardous materials into the environment. These are: (1)
an immediate cleanup or removal action if an emergency situa-
tion exists—the State is not required to provide any matching
funds;  (2) a planned removal action  (short-term response) for site
work taking less than six months and $1 million to complete; (3) a
remedial action (long-term response) for site work taking more
than six months and  $1 million to complete. Approval of the re-
quest was denied on June 14,1982, because the USEPA determined
that the sites did not meet the regulatory requirement for a planned
removal action. A  threat must exist at the site which, if left un-
mitigated, would  result in the need for an immediate  removal
action  according to the draft National Contingency Plan  of 1980.
However, that particular section was modified in the final NCP,
promulgated in 1982 and the request  was resubmitted based on the
concern  that the "public  and/or environment would be at risk
from exposure to hazardous substances if response is delayed at
a release not on the  National  Priorities List"  [National Contin-
gency Plan, 40 CFR Part 300.67 (a) (2)]. Approval of this new re-
quest was granted on July 14, 1983, and USEPA Region VIII,
the State of Utah and USDOE completed the assessment and re-
sponse actions described below.

Contracting

  During the spring  and summer of 1983,  the USEPA  and the
State of Utah negotiated a contract for the conduct of a CERCLA
Planned Removal Action with the USEPA taking the lead role. In
July, 1983, the USEPA approved the Action and  the following
September, Utah signed the USEPA/State contract, agreeing to a
 10% cost share in the project and thereby allowing it to proceed.
Utah agreed to monitor on-site activities through a local Sanitarian
attached to the District Health Department and to assist in imple-
mentation of the Community Relations Plan developed by the
USEPA.
  In the fall of 1983, initiation of the planned removal action was
delayed due to insufficient time remaining to complete the project
before the onset of winter. Monticello, on a plateau at 7000  ft
above  sea level, experiences generally harsh winters and any work
involving the pouring of concrete outdoors becomes very difficult.
Therefore, the USEPA decided to complete its arrangements for
cleanup so that construction could begin early in the spring of 1984.
These arrangements included negotiation of an Interagency Agree-
ment (IAG) with USDOE to use their expertise and contractors for
the actual cleanup work.
  In January, 1984, the USEPA concluded the IAG with the DOE;
the USDOE would act as the USEPA's prime building contractor
during the removal, overseeing  the development of the site specific
Health and Safety Plan, the contract bidding process, the mitiga-
tion activities (i.e., demolition, removal and reconstruction) and
the final decontamination certification. USDOE agreed  to carry
out these on-site activities and assigned a resident Building Inspec-
tor to oversee construction.

Health and Safety

  As part of the decision process leading to the planned removal
action, a health risk assessment was made for the occupants of both
the store and the house. The owner of the house had voluntarily re-
located his family for the winter, thus eliminating any additional
radiation exposure to them, but the operator of the store planned
to occupy it until cleanup began. The staff in the store would thus
incur additional radiation exposure during the winter months un-
less some interim measures  were  taken.  The USEPA Technical
Assistance Team (TAT) was therefore directed to investigate meas-
ures  to reduce, temporarily,  the elevated radon daughter concen-
trations in the store.
  Using information provided by previous on-site measurements,
the TAT concluded that either air cleaning  or additional ventila-
tion  could provide the desired degree of reduction. Both concepts
were evaluated for the expected radon daughter reduction (RDC),
operating costs, ease of installation and the degree of disruption of
the store personnel. The TAT  subsequently recommended the in-
stallation  of two 650 ftVmin electrostatic precipitators just below
ceiling level in the main showroom area (Fig. 2) to maximize air
circulation. The installation was completed on November 30, 1983,
and the effect on the RDC was immediately apparent.
  While this reduction was most impressive, it was only considered
as an interim measure because the efficiency of the air cleaners was
dependent on periodic cleaning and because this action had no
effect on the gamma radiation exposure rate which was a secondary
hazard. The latter could best be reduced by removal of the radio-
active material.
  The store was vacated by the lessee, at his own expense, on Mar.
10,1984. On Mar. 14, after worker health and safety briefings were
conducted by  USDOE, the building subcontractor moved on-site
and began demolition.
  The store was a rectangular (28 ft x 40 ft), slab-on-grade struc-
ture  built in 1945 and located along the Monticello main street. The
mill  tailings had been incorporated into the adobe blocks which
formed the structure walls and an old chimney. They were also used
as backfill under the floor slab (Fig. 4). The only uncontaminated
parts of the structure were the roof and the footings. During demo-
lition,  small areas of contamination were also found  under the
floor of a rear storeroom and under the asphalt of the parking lot.
Structures located immediately adjacent to the store were built at a
later date and were not contaminated.
  The entire  structure, except for the footings, was demolished
and  the contaminated material was removed to a nearby USDOE-
controlled tailings disposal site south of town by March 22. After
removing the  floor slab, approximately 16 in.  of contaminated
backfill were excavated. In all, 130 yd3 of material were removed.
USDOE and the Utah Bureau of Radiation Control monitored the
action to ensure that all contaminated materials were removed be-
fore allowing  reconstruction to begin. After a final inspection,
the leasee resumed occupancy on May 20, 1984. Subcontract costs
for rehabilitation of the structure were $73,000.
   The house, built in 1944, is a roughly rectangular (35  ft x 41 ft),
two-story structure with a full basement situated on a 125 ft x 214
ft landscaped  lot. During construction,  mill tailings were incor-
porated into the adobe blocks comprising the main floor exterior
walls and the central chimney. Radioactive tailings were also used
as backfill around the basement walls and to fill in depressions
around the lot (Fig. 5).  Preliminary readings, taken during the
engineering assessment, indicated that tailings were as deep as 4 ft
around the foundation. The foundation  itself, the upper story of
frame construction above the adobe and the roof were uncontam-
inated.
   Under provisions of CERCLA and the NCP, the USEPA may
remove contaminants during an authorized Removal Action and
repair any damages caused by or during the removal, restoring the
property as nearly as possible to the original condition.  Replace-
ment or relocation,  however,  is not permitted  except under the
Remedial Program provisions of CERCLA and  was thus  not
possible for this Removal. In this instance, the property owner re-
quested  that  the USEPA  completely  demolish the 40-year-old
house and replace it with a smaller, slab-on-grade house sufficient
to his current needs.  He felt that this would be  a  less costly
                                                                                              SITE REMEDIATION
                                                          507

-------
approach and therefore a more reasonable utilization of CERCLA
funds. However, the Removal  provisions of CERCLA preclude
any substantial modification to original  designs over and  above
those necessary to upgrade the swelling to meet current building
codes.
  The actual Removal  involved two  separate but related opera-
tions: (1) rehabilitation  of the structure, and (2) rehabilitation of
the lot. Following a health and safety briefing for the building con-
tractor by USDOE on June 6, 1984, the work crew moved on-site
and began removing the house contents to a temporary storage
area.  After the contents had been  stored, a frame support struc-
ture for one of the second story walls was fabricated and installed.
After installation, the first floor adobe walls below the  support
structure were removed and a new frame wall was built. After this
partial  demolition/reconstruction,  the  support  structure  was
moved to the next wall and the operation  was repeated—so on
until all the outside walls had been rebuilt.  In this manner, the  first
floor ceilings, second story walls, and the roof were preserved in-
tact.  During this same  period,  the adobe chimney was removed
and replaced with new brick.
  As soon as the major construction on the house was completed
to the point where it could be sealed  against recontamination by
airborne dust and dirt, excavation of the tailings from areas around
the yard began. In all,  approximately 1200 yd'  of contaminated
material were removed to the nearby USDOE disposal site.  Clean
soil was brought to  the lot and used  to backfill the depressions.
Smaller shrubs  were removed, as necessary, along with the exca-
vated tailings and replaced with local nursery stock. Where pos-
sible, tailings around the larger trees were hand-excavated in order
to preserve the trees. As before, the USDOE and the Utah Bureau
of Radiation Control were on hand during the tailings removal to
monitor the action and ensure that all contaminated materials were
completely removed. The entire lot was then covered with new sod.
                                                        Subcontract costs for the structure rehabilitation and tailings re-
                                                        moval from the yard approached $129,000 when the removal was
                                                        completed in mid-September 1984.

                                                        CONCLUSIONS

                                                          Elevated levels of radon daughter products were discovered in
                                                        two  structures in Monticello,  Utah. The radiation  source was
                                                        radioactive mill tailings  sand (a waste residue from ore extraction
                                                        processes) used as fill and building material when the structures, a
                                                        house and a store, were built during the mid-I940s. The tailings,
                                                        which contained small amounts of uranium and radium, originated
                                                        from a local World War Il-era vanadium processing mill. USDOE,
                                                        present-day custodian of the  millsite. determined that it did not
                                                        have the legal  authority to address the health  threat to the occu-
                                                        pants of the structures, the USEPA undertook a Superfund Plan-
                                                        ned Removal Action to eliminate the exposure source and the asso-
                                                        ciated health threat.

                                                        REFERENCES

                                                        I. American Cancer Society, Cancer Facts and Figures. 1984.
                                                        2. Cohen,  B.L., "Health  Effects  of Radon from Insulation of Build-
                                                          ings". Health Physics 39: 1980, 937-941.
                                                        3. U.S.  Environmental Protection Agency,  Indoor Radiation Exposure
                                                          due to Ra-226 in Florida Phosphate Lands. EPA  520/4-78-013,  Feb.
                                                          1979.
                                                        4. U.S.  Environmental Protection  Agency, Final Environmental Impact
                                                          Statement for Remedial A ction  Standards for Inactive Uranium  Pro-
                                                          cessing Sites (40CFR 192), EPA 520/4-82-013-1, Oct. 1982.
                                                        5. Federal  Radiation Council Staff Report No. 1, Background Material
                                                          for the Development of Radiation Protection Standards, May 1960.
                                                        6. Utah  State Historical Society, San Juan County, Utah,  Allen Kent
                                                          Powell (ed.),  1983.
508
SITE REMEDIATION

-------
            BIDDING CONSIDERATIONS  FOR SUPERFUND
                                    CLEANUP CONTRACTS

                                           WILLIAM F. BONNEAU
                                              ROBERT  F. SMART
                                         U.S. Army Corps of Engineers
                                                  Omaha District
                                                Omaha,  Nebraska
INTRODUCTION

  In the last 2 years, procedural errors by contractors submitting
bids for Superfund cleanup contracts advertised by the U.S. Army
Corps of Engineers (Corps) have occurred frequently. Often these
errors have resulted in the disqualification of otherwise capable
contractors. Bid documents from past cleanup projects have been
reviewed and the procedural problems analyzed. This paper is in-
tended to clarify the Corps procurement method and identify com-
mon or recurrent bidding errors.

BACKGROUND
  The USEPA and the U.S. Army Corps of Engineers entered into
an Interagency Agreement (IAG) on Feb. 3, 1982. The IAG iden-
tifies those activities for which the Corps will be responsible. One
such assigned activity is the procurement of construction contracts
for the cleanup of Federal Lead Superfund Sites. Wording in the
IAG allows the Corps to utilize existing procurement methods for
design and construction contracts. These contracting methods are
based on the Federal Acquisition Regulations and are supported by
years of experience.
  The Corps has chosen a design center approach for the contract-
ing, review and coordination of project design. Nationwide, the
Missouri River  Division (MRD) has been assigned responsibilities
for all activities from technical assistance during  the remedial in-
vestigation/feasibility study phase to  award of the construction
contract. Therefore, MRD, through its district offices at Kansas
City and Omaha, retains responsibility for a project during the pro-
curement phase. The Kansas City District is the  lead district for
USEPA Regions II, IV, VI, VII and X. Omaha is lead district for
Regions I, III, V, VIII and IX.
  The method most commonly used by the Corps  of Engineers for
cleanup activities is formally advertised competitive bidding.  This
method requires that the contractor submit a firm bid price to per-
form all work identified by the plans  and specifications  accom-
panying the Invitation for Bids. In rare cases, a Request for  Pro-
posals under a negotiated competitive  procedure may be used.
Under the  negotiated procedure, prospective contractors are re-
quired to submit a proposal outlining the methods and approaches
by which they will comply with the  specifications. The evaluation
and award  procedure for negotiated contracts differs greatly from
that of the advertised procurement but will not be discussed here
because of the projected limited use of the negotiated procedure.
The contract format used will normally be a standard construction
contract containing standard construction General Clauses, Special
Clauses, Davis-Bacon rates, etc. A service contract format may be
used when the majority of the work to be performed is other  than
that considered to be construction-related. While the standard bid
forms, General Clauses, and Specifications format may differ from
the construction format, the bid requirements for both service and
construction contracts are very similar.
BID PROCEDURES

  As indicated above, the method most commonly used by the
Corps of Engineers is a construction contract for advertised com-
petitive bidding. The procedure begins with the advertisement of
the proposed project in the Commerce Business Daily, commonly
30 days prior to bid opening. In addition, an advance notice of the
proposed contract is mailed  to a list of prospective bidders who
have expressed an interest in the specific type of work. For Super-
fund projects, a separate list of prospective bidders will be main-
tained by the Kansas City and the Omaha Districts. The Invitation
for Bids establishes a place,  date and time at which a public bid
opening will be conducted. In order for a bid to be considered for
award, it must be received at the place designated prior to the date
and time established for the bid opening.  The only exception relates
to mishandling of a bid sent by Registered Mail at least 5 days in
advance of the opening date. The instructions in the Invitation for
Bids specifically detail the application of this exception.
  After  public opening, the  apparent low bid is reviewed by the
Corps' procurement personnel. The design district retains respon-
sibility for a project until the contract is awarded. Therefore, prep-
aration of bid documents, advertisement, amendments, bid open-
ing and award are all coordinated by one Corps office. Bids are re-
viewed to  ascertain if they are responsive and if the bidder is re-
sponsible.  Responsive relates to  compliance with contractual re-
quirements. A change, addition or omission to the bid documents
which affects the enforceability of the contractk limits the contrac-
tor's responsibilities to less than those specified or would allow the
contractor to acept or reject the contract as he chooses, makes the
bid nonresponsive. Such errors are normally cause for rejection of
the bid. Responsibility pertains to the contractor's  ability to per-
form the work. Itsme such as the contractor's experience and letters
from disposal facilities reflect responsibility. Such  items may re-
quire further clarification during  bid review.
BIDDING INFORMATION

  Information describing bid procedures is contained in several
sections of the project specifications. These may include:
•Invitation for Bids
•Instructions to Bidders
•Bid Form
                                                                                           SITE REMEDIATION
                                                       509

-------
•Special Clauses
•Representations and Certifications
•Measurement and Payments
  Each section should be read carefully to assure that appropriate
guidance is  reflected in the bid documents as submitted. Often
errors can be avoided if bid requirements are reread and all pos-
sible information assembled well before bid opening. Unlike ques-
tions regarding the technical portion  of  the specification, which
should be submitted in writing  and responded to by amendment,
questions regarding procurement can  usually be handled quickly
and informally. In fact, a  separate contact is  often provided for
procurement related questions.

BID REVIEW

  The following discussion is intended to reflect the  more com-
mon errors made by past bidders. It is not to be construed as a com-
plete list of all possible errors.

Bid Form

•Bids must be submitted on the form  supplied with the specifica-
 tions or on a copy of that form.
•All numbers and written information must be clear and legible.
 If erasures or other changes appear  on  the forms, each erasure
 or change must be initialed by the person signing the bid. The bid
 price must be based on construction as  described in  the project
 specifications.
•The bidder must not submit a  bid based on alternative means or
 methods not provided for in the contract.
•The bid form must be complete and unaltered. Qualification of
 a bid, any addition or change to the bid form which alters contract
 requirements or affects contract price, is cause for disqualifica-
 tion.
•Failure  to acknowledge receipt  of all amendments can  result in re-
 jection of the bid.
•The bid form may require a single bid amount or it may require
 prices for more than one bid item. When  the bid  form requires the
 contractor to submit prices for all items, failure to do so will dis-
 qualify the bidder.
•When submission of a price on all items is not required, bidders
 should insert the words "no bid" in the space provided for any
 item on which no price is submitted.
•All bid forms must be manually signed.

Bid Guarantee

•Where a bid guarantee is required by the invitation  to bidders,
 failure to furnish a bid guarantee in the proper form and amount,
 by the time set for opening of bids, may  be cause for rejection of
 the bid.
•Acceptable forms of the bid guarantee  are described in the In-
 structions to Bidders.
•The amount of the guarantee,  usually described in percent of the
 total bid price, is presented in the Invitation for Bids.
•Usually, the bid bond amount may be expressed  in terms of a per-
 centage of the total bid price or expressed in dollars and cents.
•The bid guarantee must be manually signed.

Power of Attorney

•When the bid guarantee utilized is a bid bond, the bond is usually
 accompanied by a Power of Attorney, authorizing an agent to act
 on behalf of an insurance company.
•The Power of Attorney must be properly  completed and signed.
•The date must be such that it  is effective when the Bid Bond is
 signed and the bids are opened.
                                                        •The Power of Attorney must not restrict the financial authority to
                                                         an amount less than that required by the contract documents.

                                                        Additional Requirements
                                                        •The contract documents often require the bidder to submit addi-
                                                         tional information as part of the bid. The documentation required
                                                         will be described in the Invitations for Bids or the Bid Form.
                                                        •"Representations and Certifications," Standard Form 19B, is in-
                                                         cluded in the project specifications and is to be submitted with the
                                                         bid.
                                                        •Available plant, bidder qualifications and evidence of experience
                                                         are all items which the bidder may be required to document.
                                                        •The bidder may be required to include with the bid the qualifica-
                                                         tions and a letter of commitment from subcontractors respon-
                                                         sible for transportation of hazardous materials and disposal facil-
                                                         ities.
                                                        •The sections addressing these requirements should be read care-
                                                         fully to ensure that all requirements are met.

                                                        Submission of Bids

                                                        •Bids must be sealed, marked and addressed as directed in the In-
                                                         vitation for Bids.
                                                        •Information regarding withdrawal, modification and late receipt
                                                         of bids is presented in the Instructions to Bidders.
                                                        •Bids will be publicly opened at the time and place set for opening
                                                         in the Invitation for Bids.

                                                        PROTESTS
                                                         To ensure that Corps procurement actions are fair and that pro-
                                                        cedures are consistently applied, there is a protest  mechanism avail-
                                                        able to the bidders. By submitting a written protest to the office
                                                        issuing the Invitation for Bids, the contractor can bring an issue to
                                                        the Corps attention. The issue will then be reviewed and evaluated
                                                        to  ascertain its  validity. A  response is  prepared and  forwarded
                                                        through administrative channels  to the office designated to decide
                                                        the validity of the protest.
                                                         Unfortunately, the review  which follows a protest can be time
                                                        consuming and may hold up award of the contract, thereby delay-
                                                        ing initiation of cleanup activities and, since the evaluation is based
                                                        strictly upon adherence to  the  applicable laws and regulations
                                                        which  affect  the  procurement, it  is an  unsuitable forum for
                                                        addressing any issue not directly related to the action. The protest
                                                        privilege provides the bidder with recourse, but restraint must be
                                                        used to avoid abuse.

                                                        CONCLUSIONS

                                                         Current information indicates the number of Superfund cleanup
                                                        contracts awarded will increase dramatically  over  the next  few
                                                        years, and it is anticipated that Corps-procured  contracts will in-
                                                        crease proportionally. The Corps' procedures,  while not simple,
                                                        are necessary  for legal and  administrative reasons.  Most proce-
                                                        dural errors can be eliminated by careful review of the project spec-
                                                        ifications and early  preparation of those  forms and documents
                                                        which lend themselves to such action. Questions concerning pro-
                                                        curement can be answered expeditiously by a Corps contact, iden-
                                                        tified in  the Invitation for Bids.
                                                         Rejected bids  represent a loss to all involved. To the contractor,
                                                        it means lost effort in preparing the bid and a lost contract oppor-
                                                        tunity. To the Superfund program, rejected bids mean higher con-
                                                        tract prices and, too often,  delays associated with protests.  The
                                                        Corps appreciates the efforts of participating contractors and looks
                                                        forward  to continuing interest and cooperation  in accomplishing
                                                        what it considers a most important national program.
510
SITE REMEDIATION

-------
THE  U.S.  ARMY INSTALLATION  RESTORATION  PROGRAM

                                          ANDREW W. ANDERSON
                                    LT. COLONEL PAUL E. COUTURE
                             U.S. Army Toxic and Hazardous Materials Agency
                                     Aberdeen Proving Ground, Maryland
INTRODUCTION

  The  U.S.  Army  Toxic  and  Hazardous  Materials Agency
(USATHAMA) is currently conducting a long-range program to
determine what Army installations throughout the United States
are contaminated with hazardous wastes from past operations and
to initiate corrective actions if the wastes present a real or potential
hazard to man or the environment. As part  of its mission, the
Agency is also responsible for evaluating and removing, as war-
ranted, hazardous wastes on properties that are scheduled for re-
lease from the Army for use by others. The program, known as the
Army's Installation Restoration Program (IRP), has been ongoing
for the past nine years and has become the model for the Depart-
ment of Defense (DOD) program to assess and control the migra-
tion of environmental contamination that may have resulted from
present or past disposal activities.
  The base level of funding for the IRP has been in excess of the
$16 million estimate since all the identified problems have not been
quantified and the lack of criteria or standards for cleanliness re-
quires that each remedial action be negotiated with the appropriate
state and Federal regulatory authorities.

BACKGROUND
  The Army's IRP predates CERCLA by five years. Presidential
Executive Order  12316 delegated certain authority specified in
CERCLA to  the Secretary of Defense, and DOD was authorized
oversight over its own program by the Executive Order, thus sep-
arating it from CERCLA or "Superfund."
  Within DOD, the Secretary of Defense's authority in Executive
Order 12316 was redelegated to the Secretaries of the Army, Navy
and Air Force. The Assistant Secretary of Defense for Manpower,
Reserve Affairs and Logistics on Nov. 20, 1981, formally identified
DOD's functioning IRP as the DOD Superfund program.1  The
objectives of the IRP are these:
•To identify and evaluate past hazardous material disposal sites on
 DOD facilities;  to  control contamination migration that  pre-
 sents a hazard to health or welfare
•To review and decontaminate as necessary land and facilities ex-
 cess to DOD's mission
  In the Defense Environmental Quality Program Policy Memor-
andum (DEQPPM) 81-5, dated Dec. 11, 1982,2 DOD required that
the military departments and the Defense Logistics Agency estab-
lish and operate their own IRP. Two separate  but related require-
ments led to the creation of the Army's IRP in 1975. FIRST, the
Army was faced with regulatory agency action at several installa-
tions where past waste disposal practices had caused contamina-
tion of surface streams and groundwaters. In  the spring of 1974,
pollutants were found in water migrating from Rocky Mountain
Arsenal (RMA), Commerce City, Colorado. This prompted  the
Army to take corrective action and also to take a close look at past
military operations at other installations in an effort to determine
if they could be causing similar contamination problems.
  The SECOND requirement for this program was the need to  de-
contaminate Army-owned real estate that was considered excess to
Army needs. Efforts were needed to insure the property presented
no health or safety hazards upon its transfer to future users.
  The initial installation restoration charter specifically designated
RMA and Weldon Spring Chemical Plant ([WSCP] Missouri) as
priority tasks for the IRP. The contamination at RMA has been a
matter of public and Congressional concern for more than 25 years.
WSCP, on  the other hand, is part of a former Army munitions
plant that was later used by the then Atomic Energy Commission
(now the Department of Energy) for nuclear ore processing. It is no
longer needed by the Army, but cannot be released for general  use
due to radiological contamination.
  Most of these problems were not caused by neglectful manage-
ment practices or a disregard of environmental regulations. Rather,
they represent the consequences of following industrial and  waste
handling practices that were approved and common at the time.
  The  complexity and magnitude of the Army's  requirements in
these areas were also rapidly defined.  While the Army's environ-
mental problems were similar to  those in the civil sector, several
factors indicated  a need to establish an in-house technical  capa-
bility in this area.  (1) Since the Army has facilities located through-
out the United States, it was faced with contamination problems
under  an infinite  variety  of climatic and geohydrological con-
ditions. (2) Due to  the diversity  of the Army's manufacturing,
testing  and  training missions, the full spectrum of contamination
problems was being discovered.  These ranged from  surface  soil
contamination and associated runoff problems in the vicinity of
industrial sites to groundwater  contamination below active  and
former landfills. (3) Many of the contaminants were determined to
be unique to the  military's defense mission, e.g., wastes and  by-
products from munitions production or testing activities. For these
compounds, insufficient human  health  or environmental hazard
data were available to define risk levels and, therefore, identify
the extent of the necessary control effort.
  In recognition of the complexity, high cost, involvement of  a
wide range of Federal and state agencies and the overall impact of
such an undertaking, the Assistant Secretary of the Army for In-
stallations and Logistics directed that the Army's installation  res-
toration  efforts by  placed under project  management control.
This organization,  the  Department of the  Army (DA) Project
                                                                                          SITE REMEDIATION
                                                       511

-------
Manager for Chemical Demilitarization and Installation Restora-
tion, since redesignated  USATHAMA, is located  at  Aberdeen
Proving Ground, Maryland.
  The  first task was to identify both those Army  installations
known or suspected of containing contamination as a result of past
manufacturing, testing or disposal activities and also those installa-
tions earmarked for  excessing. Priority was given to installations
where contamination was or had a high potential for  migrating
beyond installation boundaries. Some 66 installations were initially
identified, and an assessment program was begun.  In  the  inter-
vening nine years, the list has grown steadily and now totals some
197 sites.
  The  primary hazardous wastes of concern on  Army properties
from past operations are residues from explosive-related opera-
tions, spent volatile organic solvents and heavy metal residues from
combat vehicle and metalworking operations.

IRP DESCRIPTION
  The basic IRP structure is outlined in DEQPPM 81-5. The con-
cept plan is divided into four specific phases: (1) installation assess-
ment, (2) confirmation, (3)  technical data base development and
(4) operations.
Installation Assessment
  Phase I of the IRP is an installation assessment study designed
to  be  completed  within  nine  months.  The  DOD  goal  under
DEQPPM 81-5 is to complete all records search reports by the end
ofFY85.
  In this phase, installation  files are examined, current employees
and key former employees are interviewed and the terrain and facil-
ities are examined. Additionally, all available information on past
mission, current operations, waste generation,  disposal and hy-
drogeology of the area is collected. Limited soil and water sam-
pling may also be conducted  to determine if contaminants are pres-
ent. A decision to proceed to the next phase comes at the end of the
records search is the  results show a potential for migration at haz-
ardous levels.
Confirmation

  Phase II is referred to as the confirmation phase. Phase II pre-
liminary survey studies are designed to be completed in one year.
If a comprehensive survey were required, the time would extend at
least another year.
  In this phase, preliminary and comprehensive  surveys are con-
ducted to fully define the problem  through environmental sam-
pling and analysis. Data are developed to fill identified  informa-
tion gaps revealed  during Phase I, and survey data from all tech-
nical areas are interpreted and interrelated. A decision point is in-
cluded after the preliminary survey to determine if  a confirma-
tion survey is needed to fully  define the migration problem.

Technical Data Base Development

  Phase III is referred to as technology base development.  In this
phase, control technology is matched  with specific contamination
problems at a given site to determine the most economical solution.
If control technologies  do  not exist, they are developed in this
phase. The length  of Phase  III studies varies significantly; one to
two years is an acceptable time for completion. In the case of re-
search and development, time may extend longer  is regulatory
agencies require or demand pilot testing prior to granting accep-
tance of a technology for treatment/disposal.
Operations Phase

  Phase IV of the IRP is the operations phase. This phase includes
design,  construction, operation of pollution abatement/control
facilities and the completion of remedial actions. This phase could
include the construction of containment facilities or decontamina-
tion processes and associated long-term monitoring systems. Phase
IV  operations range  from a matter  of months  to  multiyear re-
medial actions.
                                                        MANAGEMENT APPROACH
                                                          The engineering management approach and studies used to de-
                                                        fine contamination migration problems and to address their resolu-
                                                        tion are outlined in Figure 1.
                                                          The first step is  a "Records Search" which  is conducted on-
                                                        site by a team of engineers, chemists, a geologist and an environ-
                                                        mentalist (biologist or zoologist). This team reviews old files, in-
                                                        terviews present employees and retirees, examines the terrain and
                                                        facilities from the air and the ground and collects all available in-
                                                        formation relating to mission operations, waste generation and dis-
                                                        posal and gcohydrology of the area. These data are then compiled
                                                        and evaluated to determine the  potential for hazardous migration.
                                                        Should such potential exist, limited  investigative water and soil
                                                        sampling and analyses could be performed to validate or invali-
                                                        date that migration potential. If these limited investigations prove
                                                        positive, a thorough survey  of the installation  is made to com-
                                                        pletely define the pollution sources and pathways. The necessary
                                                        restoration technology to eliminate or contain the migrating con-
                                                        tamination is developed, piloted and put into operation.
                                                          Throughout this process, financial  and  personnel resources are
                                                        expended only at a  level necessary to permit decisions concern-
                                                        ing migration potential and  probable remedial actions. Through
                                                        this approach, the Army is developing a substantial data  base of
                                                        information and problem solving methodologies that are potential-
                                                        ly applicable not only to the  installation being studied but also to
                                                        other sites.

                                                        INSTALLATION ASSESSMENT

                                                        Records Search
                                                          The Records Search Program began in 1976.  The current  DA
                                                        IRP includes 197 installations (including 14 revaluations) assessed
                                                        by the end of FY85. By Sept. 30,  1984, the on-site visit phase was
                                                        completed.  There were  153  reports  representing  189  sites com-
                                                        pleted and published at an average cost of $39,343 per site.  A sum-
                                                        mary by major Army command is given in Table 1.  The remain-
                                                        ing five  reports  representing eight sites will be  published by the
                                                        end of FY85 in response to DOD direction.
                                                          Between 1976 and 1981, records search tasks were performed by
                                                        the Chemical Systems Laboratory, now known as the Chemical
                                                        Research and Development Center of Aberdeen  Proving Ground,
                                                        Maryland. The decision was made in 1981 to contract the remain-
                                                        ing program with private industry.
                                                          As the records search program was getting underway, the Army
                                                        learned in 1978 that it had an  unsuspected legacy in some of its
                                                        former properties. Interest by  the State of New York following
                                                        the Love Canal story led to inquiries about Lake Ontario Ordnance
                                                        Works (LOOW) in  Lewiston, New York. LOOW had produced
                                                        trinitrotoluene [TNT] during World  War II. It, and many other
                                                        facilities like it around the United States, were sold after the war.
                                                        This  inquiry and others like it caused  the Army to perform 21
                                                        archive searches (Table 2) to identify these former properties and
                                                        potential problems that might need attention.
                                                          Most plants ceased operations immediately following World War
                                                        11; some reopened during the Korean Conflict; and other  proper-
                                                        ties such as Nike sites were opened during the 1950s and 1960s to
                                                        be closed in the early 1970s.  When the doors to  these places were
                                                        shut, environmental  concerns were not an issue; disposal by burial
                                                        or dumping was an acceptable  practice and decontamination,
                                                        which was required and performed at that time, did not meet to-
                                                        day's standards.
                                                          Because some of the sites were found to be potential problems,
                                                        remedial  actions  were initiated. Work is currently underway by
                                                        USATHAMA at Phoenix Military Reservation ([PMR] Nike site in
                                                        Maryland) and West Virginia Ordnance Works (WVOW), Point
                                                        Pleasant, West  Virginia. During  1984, the Corps of Engineers
                                                        Civil  Works was tasked  to assume responsibility for all formerly
                                                        owned sites within  DOD. That program is  now  underway; how-
                                                        ever,  the prior owning service has the first option to conduct the
                                                        necessary environmental work if it so desires.
 512
SITE REMEDIATION

-------
                           laoie i
 Summary of Recorded Searches Performed by Major Army Commands
Installation Assessment Program, FY76-85
Major
Command
DARCOM
FORSCOM
TRADOC
WESTCOM
DLA
MTMC
HSC
MOW
COE
INSCOM
USMA
CC
OTHER
Initial
Assessments
85
50
25
9
7
3
3
3
4
2
2
1
2

Reevaluations
13

1










 TOTAL
                       196
                                               14
                          Table 2
 List of Archive Studies Performed on Formerly Owned Army Properties
                        Archive Studies
Arkansas Ordnance Plant
  Jacksonville, AR
Black Hills Ordnance Depot
  Igloo, SD
Chickasaw Ordnance Works
  Millington, TN
Cold Spring Battery Plant
  Cold Spring, NY
Erie Ordnance Depot
  Port Clinton, OH
Freeman Army Air Field
  Seymour, IN
Illinois Ordnance Plant
  Carbondale, IL
  (Crab Orchard)
Jeffersonville Quartermaster Depot
  Jeffersonville, IN
Kingman Army Airfield
  AZ
Kingsbury Ordnance Plant
  La Porte, IN
Lake Ontario Ordnance Works
  Lewiston, NY
Nebraska Ordnance Plant
  Mead, NE
New York Ordnance Works
  Baldwinsville, NY
Nike Batteries—General Search
Plum Brook Ordnance Works
  Sandusky, OH
Providence Defense Area Nike Batteries
  RI
Raritan Arsenal
  Metuchen, NJ
Santa Rosa Army Airfield
  CA
St. Louis Ordnance Plant
  St. Louis, MO
  (Hanley Area)
Schenectady General Depot
  Schenectady, NY
  (Guilderland)
West Virginia Ordnance Works
  Point Pleasant, WV
Confirmation Surveys

  Since 1979, Phase II surveys have included both exploratory and
confirmatory phases. The basis for this approach was to spend
only those resources needed to obtain the information necessary to
show an existing problem before proceeding with a full scale sur-
vey of the installation. The  exploratory phase generally deter-
mines if contamination is migrating—if not, no confirmatory phase
is necessary. If yes, the confirmatory phase defines  sources and
extent of migration.
  A total of 57 exploratory surveys were recommended either as a
result of the Records Search Program or by the direction of higher
headquarters. A detailed listing by installation is shown in Table 3.
So far, 28 surveys have been completed, 18 are  ongoing,  five are
scheduled for initiation during FY85,  two are in abeyance pend-
ing receipt and evaluation of data being generated by the installa-
tion and four are being performed by other than USATHAMA.
  The average cost of all surveys  was about $344,000 (Fig. 2).
These data show that 61% of the surveys cost less than $300,000
(actual average, $205,238). Inclusion of the four surveys that cost
between $300-$400K incorporates 73% of all surveys performed at
 a total average cost of $224,560. The six surveys costing greater
 than $600K were the more complex projects including Anniston
 Army Depot (Alabama), Alabama Army Ammunition Plant (Ala-
 bama), Letterkenny  Army Depot  (Pennsylvania),  Joliet Army
 Ammunition Plant (Illinois), Twin Cities Army Ammunition Plant
 (Minnesota)   and Frankford  Arsenal (Pennsylvania). RMA also
 cost in excess of $600K but is not represented.


TECHNICAL DATA BASE DEVELOPMENT

  For the most part, this phase is now combined with the confirma-
tion survey portion of the assessments phase providing a stream-
lined effort performed as  a  concurrent  rather  than sequential
effort. Projects which can currently be considered in this phase
are RMA, Anniston Army Depot, Letterkenny Army Depot, Twin
Cities Army Ammunition  Plant, Sharpe Army Depot (California),
Milan Army Ammunition  Plant (Tennessee), PMR (Maryland) and
Cornhusker Army Ammunition Plant ([CAAP] Nebraska). Pro-
jects which have gone through this phase are Redstone Arsenal
(Alabama), Pine Bluff Arsenal (Arkansas), Frankford Arsenal,
RMA and Anniston Army Depot.
  The Army has faced several problems in conducting this phase
of the investigation:
•For many of the hazardous problems encountered, there are no
 known standards or criteria.  It leaves the Army with the  "How
 clean is clean?" problem which is not easy to solve and relates to
 the next point.
•State standards and methods of operation are not homogeneous
 or  even published. Some  are  more actively pursuing  environ-
 mental problems than others. The result is that the Army has to
 negotiate with each  individual state to establish acceptable  re-
 sidual contaminant levels  and methods of  treatment  or non-
 treatment.


OPERATIONS
  The 26 installations where remedial actions have been performed
or actions are required are listed in Table 4. The  most significant
limitation on remedial action operations is the lack of working or
proven alternative technologies needed to cleanup currently iden-
tified problems caused by explosives or organic solvents. One either
digs up the contaminant and moves it somewhere else or covers it
in place. In the case of groundwater/aquifer contamination treat-
ment, no such action has proved to be a practical alternative.
  More competitive  pressure needs  to be placed on industry to
provide solutions incorporating technologies that  not only reduce
the enormous  cost of transportation but also eliminate the deferral
of the (problem or residual liability caused by burying hazardous
material in different "secure"  landfills around the country. While
the USATHAMA technology development program  (described
below) will  produce potentially usable technology, the Agency
needs it now but does not have it.

CASE HISTORIES
Anniston Army Depot
  Anniston Army Depot, located in northeast Alabama, originally
was an ammunition storage  depot. During World War II its mis-
sion expanded to include combat  equipment storage.  Over  the
years, it evolved into what is now the major tank rebuild facility
in the free world. Industrial processes at Anniston Army  Depot
led to  the production of waste chemicals, mainly degreasing sol-
vents and metal-processing sludges.
  In 1980,  samples of monitoring wells showed these hazardous
wastes had entered the groundwater and appeared to  be migrat-
ing from two  former disposal areas located within the southeast
industrial area of the Depot. USATHAMA subsequently initiated
a survey and assessment to determine the extent of hazardous con-
tamination migration and  to develop plans for abatement or treat-
ment as required. The program consisted of four tasks:
                                                                                               SITE REMEDIATION
                                                                                            513

-------
                                                            Table 3
                                               U.S. Army Survey Program Summary

INITIATED/COMPLETED
FY
7? Aberdeen PC (EA), MD
79 Alabama AAP. AL
79 Badger AAP. Ml
81 Blue Grass DA. KY
80 Camp 51 rams, DC
81 Cornhusker AAP, NŁ
80 Ft. GUI em, GA
79 Ft. Monroe, VA
80 Ft. Wlngate DA, NM
76 Frankf ord Arsenal , PA
82 Gateway AAP, MO
78 Hawthorne AAP. NV
81 Lexington DA, KY
81 Louisiana AAP, LA
80 Michigan AMP, MI
79 Milan AAP. TN
80 Navajo DA, AZ
76 Pine Bluff Arsenal , AR
60 Providence NIKE. RI
79 Redstone Arsenal, AL
80 Sacramento AD, CA
82 St. Louis ASC, IL
79 St. Louis OP, MO
80 Savanna DA, IL
81 Sharpe AD, CA
81 Tooele AD, UT
80 Umatilla DA, OR
77 Weldon Spr CP, MO

U.S.
INITIATED/COMPLETED
FY
81 ALABAMA AAP, AL
(LEASEBACK AREA)
82 ANNISTON AD, AL
79 FRANKFORD ARS, PA
82 GATEWAY AAP, MO
76 PINE BLUFF, AR
30 PROVIDENCE NIKE, RI
79 REDSTONE ARS, AL
79 ST. Louis ORD PLANT, MO
83 SANTA ROSA AAF, CA

• NOT PRESENTLY SCHEDULED

SURVEY STATUS
INITIATED/ONGOING
FY
81 Anntston AD, AL
84 Detroit Arsenal , MI
83 Ft. Belvolr, VA
84 Ft. Drum, NY
81 Indiana AAP, IN
80 Iowa AAP, IA
80 Jollet AAP, IL
81 Letterkenny DA, PA
81 Lone Star AAP. TX
82 Longhorn AAP, TI
84 Natlck (Sudbury). MA
84 Phoenix Military
Reservat ton , MO
84 Rtverbank AAP, CA
76 Rocky Mtn Ars, CO
81 Twin Cities AAP, MN
83 Vtnt H11I Farms, VA
81 Volunteer AAP, TN
84 West Virginia Ord
uks, wv








Table 4
Army Remedial Action Program Summary
INITIATED/ONGOING
FY FY
8*4 CORNHUSKER AAP, NE *
83 JOLIET AAP, IL
84 MILAN AAP, TN 87
78 ROCKY MT ARS, CO 85
86
85
86
86
85
86
85
85


INITIATION/SCHEDULED
FY
85 Bayonne MOT, NJ
85 Sunflower AAP. KS

•Survey pending review
of saapUng data fro* the
U. S. Arny Environmental
Hygiene Agency and the
State of Maryland

In Abeyance

Holston AAP, TN
Aberdeen PG (AA). MD
Ft. Meade, MD

Other

OLA - Ogden, UT
OLA - Richmond, VA
DLA - Meaphts, TN
Ft. Detrtck, MD









INITIATION/SCHEDULED

ALABAMA AAP, AL
(GSA s IND AREAS)
IOWA AAP, I A
LETTERKENNY AD, PA
LOUISIANA AAP, LA
PHOENIX NIKE, MD
SAVANNA DA, IL
SHARPE AD, CA
TWIN CITIES AAP, MN
VOLUNTEER AAP, TN
WEST VIRGINIA ORD WKS, WV
WOODBRIDGE RESEARCH
FACILITY, VA
Geotechnical Evaluation
  Task One, the geotechnical evaluation, was accomplished during
April to September 1981. Surface remote sensing techniques along
with geohydrologic investigations were employed to evaluate the
potential for groundwater contamination and its subsequent migra-
tion and movement into and  within the bedrock. These  tech-
niques successfully delineated the horizontal and vertical bound-
aries of seven buried chemical  sludge disposal trenches. In  addi-
tion, these techniques  provided insight to the direction of the
shallow  groundwater flow and  the depth to and configuration of
the bedrock. Based on  these results and USATHAMA's concern
                                                       that a potential existed for contaminant migration into the bed-
                                                       rock, the disposal of hazardous wastes in the seven chemical sludge
                                                       trenches was discontinued.

                                                      Groundwater Quality Assessment

                                                        Task Two planning was begun in August 1981, and a ground-
                                                      water quality assessment plan was prepared by USATHAMA and
                                                      provided to the appropriate regulatory authorities in October 1981.
                                                      The objectives of the plan and  subsequent assessment  were to de-
                                                      termine  the concentration, rate and extent  of migration of haz-
                                                      ardous waste or hazardous waste constituents in the groundwater.
514
SITE REMEDIATION

-------
                                                      PROGRAM  APPROACH
                                    PRELIMINARY SAMPLING & ANALYSIS
                                    GEOHYDROLOGY STUDIES

                                    CRITERIA/STANDARDS
SOURCE CONTROL STUDIES
WATER MGT STUDIES
                                                                 FINA
                                                                STANDARD DEVELOPMENT
                                                                PROCESS STUDIES
                                                                CONTAMINATION STUDIES
                                                                EIS
                                                                STATE PERMITS
                                                           Figure 1
                       USATHAMA  Approach to Determine Contamination Migration and Required Corrective Actions
The results of the assessment were published in an October 1982
status report.
  Based on these findings which located new areas of ground-
water contamination, a follow-on study involving computer sim-
ulation was completed to determine the sources of the contamina-
tion and to evaluate the feasibility of groundwater intercept and
treatment systems. The study was provided to state  and Federal
regulatory authorities in June 1984 for review and comment.

Economic and Technical Analysis
  Task Three, originally scheduled to begin in February 1982, was
brought forward to August 1981 based on the geotechnical evalua-
tion. USATHAMA's in-house expertise was utilized  to complete
this task which involved an economic and technical analysis of a
variety of alternatives for closure of the chemical sludge disposal
trenches and the old lagoon sludge pile. USATHAMA completed
the alternatives analysis ahead of schedule, and a recommended
approach for remediation was presented to the U.S. Army Material
Development and Readiness Command, now the U.S. Army Ma-
terial Command (Alexandria, Virginia), in  November 1981 along
with a request for funding. This approach, incorporating the most
feasible technical and economic alternative, required the exhuma-
tion of the chemical sludge disposal trenches and other wastes for
transportation and off-site disposal at a permitted disposal/treat-
ment facility located in Emelle, Alabama.
               SURVEY  COST -  MODE
    35
    30
    25
    20
    15
    10
      0   100   200   300   400   500  600  700  800  900  1,000
                         DOLLARS RANGE
                           Figure 2
        Mode Distribution of Exploratory Survey Costs ($000)
Closure
  Task Four, the closure operations, began with the preparation
of closure specifications in March 1982. Approval of these specifi-
cations was granted by the regulatory authorities in June 1981. Ex-
cavation of the hazardous wastes commenced in November  1982
and was  completed in May 1983, removing 60,000 tons of waste
and surrounding soil contaminated with spent organic solvents and
heavy metals.
Cornhusker Ammunition Plant
  Cornhusker Army Ammunition Plant was constructed in 1942 in
Hall County, 4.0 miles west of Grand Island, Nebraska. Its primary
mission was to load, assemble and pack a variety of conventional
munitions  containing  TNT and  cyclotrimethylenetrinitramine
[RDX].  Industrial activity at CAAP was  intermittent, depend-
ing on levels of military activity. Active production periods coin-
cided with World  War  II, the Korean Conflict and the Vietnam
Conflict. CAAP is currently in  a standby status. Industrial waste
generated at CAAP consisted primarily of washdown waters (con-
taining TNT and RDX)  resulting from loading and packing opera-
tions.  These wastes  were treated by  means of filters, cesspools
(dry wells) and leaching pits. CAAP did not discharge wastes to
surface waterways. Liquid explosive wastes from past operations
that were disposed into the cesspools  and leaching pits at  CAAP
contaminated the shallow groundwater in a one square-mile area
within the Plant's northeast boundary.
  The survey revealed that the soils  at the bottom of leaching pits
used to collect ammunition manufacturing wastes during World
War II and again during the Korean and Vietnam Conflicts are
contaminated to a depth of three feet or greater with explosives;
TNT,  RDX, dinitrotoluene  [DNT], their manufacturing by-pro-
ducts and their environmental  degradation products.  These con-
taminants leached  into the underlying sand  and gravel which con-
tains the shallow aquifer.
   Subsequently, results of the Army's  groundwater  monitoring
program at CAAP confirmed  the  presence of explosives in the
groundwater three miles  beyond the Plant's eastern boundary.
Chemical analyses of the 467 private wells sampled east of the in-
stallation detected the explosive compounds RDX, DNT and TNT
in two wells within one mile of the Plant. However, RDX alone was
found in 246 wells located up to three miles east of the installation
boundary. At present,  no regulatory standards have been estab-
lished for any of these compounds. The Army,  however, has
                                                                                               SITE REMEDIATION
                                                          515

-------
recommended interim water quality criteria for TNT and RDX
based on toxicity studies which  concur with the National Acad-
emy of Sciences.  The USEPA has recommended a  water  quality
criterion for DNT.
  The Army is providing bottled water for drinking and cooking
to those residents whose wells are affected as determined by local
health authorities in conjunction with the Army. On July 31, 1984,
the Army entered into contract with the City of Grand Island to
extend the city water supply to the affected community at a cost
of  approximately $5  million  dollars.  In  conjunction with  the
immediate measures needed to provide drinking water, other work
to contain or abate the sources and contamination plume  is con-
tinuing.

Frankford Arsenal

  The Frankford Arsenal (Philadelphia, Pennsylvania) dates back
to  1816. During its 161 years of operation a variety of activities
including munitions manufacture, materials and research develop-
ment activities, development of propellant and  cartridge-actuated
devices and a variety of procurement missions  were  accomplished
at the Arsenal. In 1976, the facility was declared excess to ARmy
needs and plans  were  put in place for the decontamination and
cleanup of the Arsenal prior to transfer of the property to the Gen-
eral Services Administration (GSA) for subsequent disposition and
release for unrestricted use.
   In the spring and summer of 1978, a survey of the 110-acre
Arsenal was conducted under contract. This survey identified low
levels of: (1) heavy metals residues, (2) explosive residues and (3)
radiological contaminants. Based on the results of this survey, bids
were requested from various industrial contractors to perform de-
contamination and cleanup of the  Arsenal. In  September 1979, a
contract for  the Arsenal  decontamination  and  cleanup  was
awarded.
  The cleanup program was conducted in three phases. During
Phase I, decontamination methods and procedures were verified.
Detailed standing operating procedures  were prepared  during
Phase II, and decontamination operations were conducted during
Phase III.  The contractor completed decontamination operations
in November 1980. The Arsenal was subsequently sold by GSA in
1983.

INSTALLATION RESTORATION RESEARCH
AND DEVELOPMENT TECHNOLOGY

  Concurrent with the ongoing survey and restoration efforts, a
technology development program has been initiated to obtain new
and better detection and analytical techniques and containment
and treatment processes. The areas being investigated include ecol-
ogy, geotechnicai studies, analytical systems  technology,  decon-
tamination technology and management and retrieval of the myriad
of data generated throughout the entire IRP. A synopsis  of each
area of efforts follows:

Ecology

   In ecology, technology capable of defining pathways of migrat-
ing  contaminants as determined by  their  efforts on selected en-
vironmental communities is being  applied. Aerial infrared photo
interpretation is being employed for establishing baselines for vege-
tation stress to determine the effectiveness of decontamination/
abatement measures for detecting potential sites for surface  soil
contamination and surface water migration as well as locations of
landfills, burial sites and ditches.

Geotechnicai

  Geotechnicai studies are underway to determine  the nature of
soils,  bedrock, sediment, surface  water,  groundwater and their
relationships to contaminant occurrence and movement. Develop-
ment of a surface water and groundwater quality monitoring pro-
gram involves locating stream monitor stations and drilling wells
at positions which are indicative  of the  general groundwater situa-
tion.
                                                        Analytical
                                                          Due to  the varied nature of installation restoration analytical
                                                        requirements, the analytical methods program develops or adapts
                                                        analytical  procedures for particular needs in the parts per billion
                                                        or parts per trillion range and implements these procedures in those
                                                        laboratories conducting analyses for the Army.  In some instances
                                                        where the presence  of only a few contaminants was anticipated
                                                        based on the installation's past operations, environmental changes
                                                        have transformed the compounds into potentially several hundred
                                                        distinct components. Rapid and sensitive  screening techniques are
                                                        required to determine the qualitative presence or absence in a sam-
                                                        ple. New technology to meet these ever changing challenges is con-
                                                        sistently being studied in research and development projects.

                                                        Decontamination

                                                          Many of the contaminants of concern are unique to the military,
                                                        and large  areas of contaminated  soil, sediment and surface and
                                                        groundwater represent technically complex decontamination prob-
                                                        lems and high treatment  costs. Studies are underway to  identify
                                                        and develop novel and cost effective decontamination technologies
                                                        in these areas.
                                                        Computerization

                                                          A network of computer terminals located at  field installations
                                                        has been connected  to a main-frame Univac    1108 computer at
                                                        Aberdeen  Proving Ground where  USATHAMA,  the centralized
                                                        management agency, is  located. Data  are collected through  the
                                                        terminals with mass storage of the data in the Univac.
                                                          Chemical, ecological and geological data are contour-plotted by
                                                        the computer on digitized maps  of field installations. Contam-
                                                        ination plumes  of various contaminants, in  various media (soil
                                                        surface water, groundwater, sediment and biological tissues) can
                                                        be studied over time periods to determine location and direction of
                                                        movement. A cost/benefit/risk analysis program  for decision
                                                        making in containment and/or treatment alternatives is being inte-
                                                        grated into the simulation effort.
                                                          To date, approximately three million records have  been received
                                                        and automated  on the computer. Software has been acquired or
                                                        developed for plotting chemical plumes, species demography and
                                                        some geological parameters; other programs have been developed
                                                        for food chain presentation and mass flux calculations.

                                                        CONCLUSIONS

                                                          In this paper, the authors have briefly described the Army's IRP
                                                        and some of its achievements to date. With over nine year's exper-
                                                        ience, the  Army is confident that its technical approach and cen-
                                                        tralized management structure offer an effective mechanism to deal
                                                        with the unique military environmental contamination problems.
                                                        Recent discussions with the USEPA have revealed that the Army
                                                        program conforms to and is very compatible  with parallel efforts
                                                        in the civil sector to promote remedial actions at "uncontrolled"
                                                        waste disposal sites. Further,  the Army's program has been used
                                                        as a model by the Navy and Air Force.
                                                          The key to the Army's program effectiveness lies  in its central-
                                                        ized and flexible  management structure.  With a single organiza-
                                                        tion fully informed and technically responsible for dealing with the
                                                        Army's installation  restoration problems, technological  advances
                                                        can be rapidly applied to corrective projects,  and program prior-
                                                        ities can be easily adjusted in response to newly discovered prob-
                                                        lems or changing regulatory agency requirements.

                                                        REFERENCES

                                                        1.  Information Paper on the Department of Defense Installation Restora-
                                                           tion Program, Office of the Assistant Secretary of Defense, Manpower,
                                                           Reserve Affairs, and Logistics, Apr. 4, 1983.
                                                        2.  Defense Environmental Quality Program  Policy Memorandum  81-5,
                                                           Subject: DOD Installation Restoration Program, Office  of the Assistant
                                                           Secretary of Defense, Manpower, Reserve Affairs, and Logistics, Dec.
                                                           11, 1981.
516
SITE REMEDIATION

-------
      THE  CHALLENGES OF SITING  HAZARDOUS  WASTE
                               MANAGEMENT FACILITIES

                                                  JOE TELLER
                    National Association of Local Governments on Hazardous Wastes
                                                Washington, D.C.
                                                         and
                                              Gulf Coast Authority
                                                  Houston,  Texas
 INTRODUCTION

  In this paper, an effort to site a hazardous waste disposal facility
 by the Gulf Coast Waste Disposal Authority—a description  of
 what was done and the reaction to those efforts.  Included is  an
 evaluation of why the effort has encountered so many delays and a
 prediction of what will succeed in the task of securing a site for a
 hazardous waste facility.
 Authority's Powers

  To fully understand the total situation, an explanation of several
 key components is necessary. One of these is the nature of the
 several parties involved. The Gulf Coast Waste Disposal Authority
 (the Authority) was formed by an act of the Texas Legislature in
 1969. At that time, the Houston Ship Channel was referred to as
 the most  heavily polluted body of water in the world, and may
 actually have been in the United States' top ten most polluted.
  Recognizing that the more than 150  existing governmental en-
 tities were improperly equipped to attack the broad, complex issues
 which had created the pollution problems in the heavily industrial-
 ized Houston/Galveston  Bay area, the Texas  Legislature drew
 heavily on the experience of the Ruhr  River Valley Authority in
 Germany and provided the  Authority  with unusual powers  to
 attack the water quality management problem of the Houston Ship
 Channel and Galveston Bay. Those powers include the authority to
 issue tax-exempt bonds; to construct, purchase and operate waste
 treatment facilities; to condemn land and  rights-of-way; and to
 purchase equipment and supplies on the same tax-free basis as any
 other governmental entity.
 Governing Board

  A local unit of government, the Authority has nine directors
 comprising the governing board. Consisting of three members from
 each of the three counties comprising the Authority's area of juris-
 diction, board members serve two year, overlapping terms.  From
 each county, one member is appointed by the Governor, one by the
 County Commissioners and one by a consortium of the mayors of
 all cities within that county.
  The Board sets general policy and acts in  its legislative capacity
 to pass resolutions, issue bonds and award certain contracts. The
 Board also hires the General Manager, who  directs the implemen-
 tation  of Board policy. The  Authority was initially funded by a
 grant from the state but became financially self-sufficient in 1976.
 All Authority funds now come from charges for services provided.
 In brief, the Authority is a government that operates like a busi-
ness.
Program

  Early in the Authority's existence, it was determined that, with
no less than three regulatory agencies active in the area, sufficient
regulatory activities were being undertaken. The Authority, there-
fore, set its goals toward development of operational and manager-
ial solutions to the various water quality problems of the area.
Throughout its existence, the Authority has promoted a coopera-
tive effort between the public and private sectors as the most effec-
tive approach to what are actually mutual problems.
  The Authority has utilized the regional concept, joint or com-
bined treatment facilities and operation of individual facilities, and
is currently operating 26 wastewater treatment facilities and an in-
dustrial solid waste disposal facility operated for four industries in
Texas City. The wastewater treatment facilities range  from a  55
million gal/day combined  system to a 1200 gal/day single plant.
About 90 million gal/day are treated for 66 separate entities.

HAZARDOUS WASTE FACILITY
  In 1977, the results of ongoing water  quality management re-
views indicated the need for a hazardous waste disposal facility to
serve  the Metropolitan Houston  industrial  community.  The
Authority staff conducted a very brief and specific purpose market
survey and found the types of wastes which would need to be dis-
posed of and their approximate quantities. During ensuing months,
a concept evolved which would provide an environmentally accept-
able disposal arrangement utilizing those physical  facilities which
were considered appropriate for the type waste produced in the
general Houston area. The development of that concept included
investigations of  processes,  physical equipment,  corporate pol-
icies, financial capability,  marketing  practices and  operational
records of several separate entities.
  Discussions were held with the regulatory agencies having juris-
diction over the various processes and the companies operating
physical facilities,  and contact was made with individuals having
an interest in the operations. From  these investigations, the spe-
cific processes and companies were chosen to be a part of the plan-
ned hazardous waste disposal facility.  Initial contracts were then
negotiated with the selected entities and the concept was finalized.

FACILITY CONCEPT

  That concept called  for a three party arrangement structured
thusly: The Gulf Coast Waste Disposal Authority would own the
land, hold the permits and be responsible for internal quality con-
trol, any waters discharged off-site and long-term site management.
The incinerator contractor would construct and  operate the  in-
                                                                                           STATE PROGRAMS
                                                        517

-------
cineration process including pretreatment and  resource recovery
equipment for essentially organic waste materials. The chemical
fixation contractor would build and operate the chemical fixation
process including pretreatment, resource recovery devices  for the
inorganic waste  materials and all ash and scrubber sludges from
the incineration phase.
  There are obviously many very significant items not included in
that  brief description of the  basic concept, some of which  were
included in the permit application and some of which were to be
incorporated in the contractually required operating plan. One key
point contained in the contracts and permit application  was the
obligation and right of the Authority to require the shut-down of
any process unit which was causing a violation of any condition of
any permit.
  To facilitate that control, the Authority required that all utility
services (electricity, water, etc.) would be the responsibility of and
under the control of the Authority. Under this arrangement, any
utility could be turned off until the process  malfunction was cor-
rected. Also, the central laboratory facility  would  be  owned  and
operated by the Authority, insuring no interference with the quality
control responsibility of the Authority.

Waste Treatment Processes
  The physical facilities to be included were two rotary kiln incin-
erators, one liquid combustion device,  extensive pretreatment and
product recovery facilities, waste materials storage, chemical fix-
ation process equipment, a fixation product deposition area,  lab-
oratory, administration, maintenance and related facilities and util-
ity control facilities. These were laid out to allow staged construc-
tion of a planned design of 520,000 tons/yr of waste materials.

SITE SELECTION

  Having agreed on the basic relationships  and physical facility,
the next activity was site  selection. With some input from  the pro-
cessors, the technical staff of the Authority set basic  site criteria
and  then utilized an area elimination process to narrow down the
possible list of sites. One site criterion was that the elevation of the
site be not less than 25 ft above mean sea level. Large land areas
were eliminated  from further consideration by application of that
criterion.
  Some other criteria  were proximity  to population,  distance to
major transportation routes, general subsurface soil characteristics
and  proximity to the industrial community. Further application of
the criteria eliminated even more land areas.
  Ultimately, 20 potential sites were acceptable under the given
criteria. Specific investigation of those sites was conducted, and a
site was chosen as being  optimum for the proposed location. The
Authority felt that this site offered a bonus; the property had been
used for extraction of surface deposits of sand and had large, deep,
unsightly excavations remaining. Local developers and real estate
agents assured the Authority that it was not economical to fill those
excavations and build any structures, so the concept that the ex-
cavations ultimately  would be filled  with  a stable, inert material
and  covered  with clay and then native topsoil appeared to be a
logical reclamation project. The Authority secured an option on
the site to allow site investigation and to assure that, if a permit
were authorized, the real estate would be available for purchase.

Fatal Flaw Evaluation
  The Authority's next effort was what it termed a "fatal flaw"
evaluation. Discussions with the state and federal regulatory agen-
cies  revealed those technical  site considerations considered  most
significant, and the fatal flaw evaluation was conducted to insure
that the site was acceptable from those technical considerations.
  This site did meet the requirements set forth—the man who had
extracted the sand had stopped removing sand when  he  had run
into  clay. Subsurface borings indicated that the clay  had a min-
imum  depth of 60 ft and was impervious to the extent that it met
the regulatory agency requirements with ease. In effect, the site was
a series of large, 35 to 45 ft deep holes whose bottoms and sides
were impervious clay.
                                                          When the fatal flaw evaluation proved that the site had no appar-
                                                       ent  technical deficiencies,  the  investigations  were extended  to
                                                       collect the data necessary to prepare the permit application. That
                                                       required information has grown in past  years to  become a some-
                                                       what massive array including archaeological,  terrestrial biological,
                                                       aquatic biological, detailed subsurface geological  and almost end-
                                                       less evaluation of factors relative to the site.
                                                          For example, in 1979, when this effort was underway, one of the
                                                       committee advisors pointed out that one local member of the en-
                                                       dangered species list was the Houston Toad, and the Houston Toad
                                                       might be found on this site. It was determined  that  a prudent course
                                                       of action would require an evaluation of the presence (or absence)
                                                       of the Houston Toad. That evaluation is not as simple as might be
                                                       expected.
                                                          To start,  the Houston Toad  is  almost identical in physical
                                                       appearance to the common toad,  and eyen the trained eye is fre-
                                                       quently unable to detect the differences.  The toad does not occur
                                                       in large numbers at any location, probably because of a natural in-
                                                       stinct which prompts the toad to try to mate with  an appropriately
                                                       shaped rock or discarded apple core. But, it is this mating instinct
                                                       which proved to be the key to determination of the toad's presence;
                                                       the Houston Toad emits a distinctive mating call which is identi-
                                                       fiable by a well-trained terrestrial biologist. Unfortunately, the
                                                       Houston Toad only mates during a two week period in early spring
                                                       and, even more unfortunately, that two week  period is determined
                                                       by temperature, rainfall and other less precise factors.
                                                          Consequently, to make the evaluation, one has to hire an indi-
                                                       vidual with a highly-trained, especially sensitive ear to stand around
                                                       in the rain (at night) for about a month to listen to frogs croak. As
                                                       it turned out, from all the information gathered,  there were none
                                                       present. But the Houston Toad is present in other parts of the state
                                                       rather remote from Houston. To  further compound the Author-
                                                       ity's frustration, the Houston Toad has now been removed from
                                                       the endangered species list.

                                                       PERMITTING HURDLES
                                                          This exercise has been  cited to make the point that some regula-
                                                       tory agency requirements, imposed by totally  unwarranted legisla-
                                                       tion  from that  fantasy-land on the Potomac River,  detract  from
                                                       the real intent of environmental improvement  and  add cost to what
                                                       is already a very expensive effort—the preparation  of an acceptable
                                                       permit application. The country  may already have reached the
                                                       point where the cost of preparing a permit application, not even
                                                       including the public hearings or legal expenses associated with the
                                                       process, will preclude some competent entities from attempting to
                                                       secure a permit for  needed, technically sound disposal facilities.
                                                       They simply will not have the financial resources to risk an attempt.
                                                       The  inevitable result is fewer facilities to dispose of larger quan-
                                                       tities of wastes.
                                                          One of the elements in the permit application is site design, in-
                                                       cluding a site layout showing relatively firm locations of all physi-
                                                       cal improvements. When the  Authority's investigations and data
                                                       collection had progressed to this point, it developed a public in-
                                                       formation packet and prepared a public information program com-
                                                       plete with a variety of graphic displays and technical information.
                                                          The Authority staff initiated the public information program by
                                                       meeting with a  group of area environmental organization repre-
                                                       sentatives including  the  Sierra Club, the Audubon  Society, Cit-
                                                       izen's Environmental  Coalition  and  Houston/Galveston Toxic
                                                       Substances Task Force. These representatives were present at the
                                                       first program and the response was clear and direct: the need for
                                                       good hazardous waste disposal was  present;  the  high technology
                                                       proposed was good; the location appeared acceptable—go to it!
                                                          This encouragement from a blue ribbon panel of knowledgeable,
                                                       concerned  citizens gave the Authority a  sense of satisfaction and
                                                       comfort which  later proved unfounded, but at the time it felt that
                                                       the need and technical competence of both the processes and the
                                                       site would be sufficient for the project to go forward smoothly and
                                                       quickly.
 518
STATE PROGRAMS

-------
  The Authority then met with elected officials; mayors, state rep-
resentatives and state senators, congressmen from the area and the
staffs of the U.S. Senators. Again, it received the go signal and
comments about the obvious need for a destruction process.
  A special meeting focusing on technology was held with the Pub-
lic Works Department of the City of Houston, in whose corporate
boundaries the site was located.  Again, the comments were:  the
need is desperate; the technology is sound; proceed. These munic-
ipal representatives pointed out very explicitly that illicit discharges
of hazardous waste materials to the municipal sewer collection sys-
tem were very costly to the city and resulted in water quality prob-
lems both with the treatment processes and the collection systems.

Public Information Program

  The public information program for area citizens was carried out
in two concurrent efforts: Area civic organizations were contacted
and arrangements made to  present a program at their scheduled
meetings. These civic groups included Rotary, Lions and Kiwanis
clubs; professional groups  such  as the Association of  Certified
Public Accountants; engineers' associations; chambers of com-
merce; and various other organized groups.
  The other effort involved written invitations to homeowners and
businesses in the area of the site.  A  convenient location was
arranged and evening meetings were scheduled to accommodate
residents' work schedules. The program was arranged to answer all
questions posed.
  Over a 14 month period, 56 meetings were held at  which over
1500 individuals were directly  told about the plans for the project.
Additionally, information packets were mailed or hand-delivered
to about  500 individuals or  entities.  During  these "public in-
formation" sessions, the first voices of disapproval were heard and
they were of one theme: "It sounds great and we know  it's needed,
but don't put it here. Put it somewhere else.''

Opposition Develops
  On the morning following  one evening meeting, a call was re-
ceived from a real estate agent who had been in attendance the
night before. He stated that he had available for sale a tract of land
which he considered more suitable than the  site the Authority was
proposing. Over the phone, the  staff determined the  location of
that land and informed him that it was closer to a highly popu-
lated area (the City of LaPorte) than the site selected near the City
of Pasadena where he lived. His response was, "Yes, I know that,
but I don't live in LaPorte." This message, "Not in My Back-
yard," was the only thought expressed and was coming from a
very small segment of the large group with whom we had made
contact.
  In June, 1980, the Authority formally filed the application for
permit with the regulatory agencies.  The filing set in  motion the
normal procedures, the first of which was formal public notifica-
tion of the application having been filed and setting a date for a
public hearing.  The hearing was  held  at San Jacinto College in
Pasadena, near the proposed  site, about 30 days after the public
notice appeared. During this period, a group of residents, located
approximately two miles from the  site, organized themselves in
opposition to the  project.  They called their group the  Citizens
Against Polluting Our Neighborhoods  Environment (CAPONE).
These citizens acknowledged  the harsh implications of the name
and stated they felt it would take forceful measures to impact the
proposed facility.
  Curiously, this small neighborhood group was unusually well fi-
nanced, being able to commission a public relations firm to pre-
pare a 16mm sound and  color film, print three color pamphlets
and retain an attorney. The Authority learned some time later that
the financing for their effort was  provided by an existing waste
disposal company  which  apparently did not wish to  show overt
opposition. They spent in excess  of $100,000 in opposing, by sev-
eral methods, the permit application.
  The Authority attempted to meet with the CAPONE group on
several occasions but was  finally informed that the  opposition
group did not wish to secure any facts about the project; they were
determined to oppose the location. CAPONE was not alone. They
were joined by another area self-styled environmental  group and
numerous other cities and local entities to oppose the project.
  At  the first  official  public  hearing called by the  regulatory
agency, the attorney representing many of the opposing groups
argued that he had not had sufficient time to prepare his case. He
asked for and received a postponement. That was the  first lesson in
what is termed "legalistic delay." Over the next several months,
additional lessons were learned.

City Opposition

  The first real surprise came when the City of Houston  announced
its opposition to the site.  Having met with the Public Works De-
partment and the Mayor's office, the  Authority was naive in  its
assumption that all was well. The Authority was unable, at  that
time,  to learn of the basis for opposition. Now came a most frus-
trating period which saw legal postponements of scheduled depo-
sitions, the unusual circumstances concerning the withdrawal  by
the Authority's  legal firm from representation in this matter and
the false and misleading stories being told about various aspects of
the project.
  During this period, a representative of the City of Houston ad-
vised  the Authority's new attorney that if the project would be re-
located to an industrial area, then the City of Houston would with-
draw  its opposition. If the  Authority  persisted  in its application
at the chosen site, the City would utilize all avenues of opposition,
including appeals to any issued permit.
  In  considering its position, the fact that the City maintains a
large  contingent of attorneys available to  prepare and litigate
appeals was most significant. The Authority, therefore, elected to
withdraw the application and regroup for another application at a
site within an industrial district.
  In attempting to understand the position taken by the City  of
Houston with respect to the application, the Authority inquired
most thoroughly of representatives of the public works department
of the City of Houston, and was advised that it should contact the
mayor's office for an explanation. The public works department
people stated  that they were directed  by  the mayor's office  to
develop  a technical basis  for opposition to the application.  The
most  logical point to cite  as  the reason permits  were not  secured
for the facility is the opposition of the City of Houston.  The City's
ability to delay ultimate completion of the project for an extended
period of time would certainly cast doubts on the economic  via-
bility  of the project. The Authority does not yet know precisely
what  triggered the  obviously politically based opposition of the
City.
Depth of Opposition

  The preceding discussion  has  described the basic activities  and
actions relative to the aborted effort to secure a  permit for a high
technology hazardous waste  disposal facility. The discussion did
not detail the extent of the activities of those in  opposition to the
application. It is not that one should be  unaware of those activities,
but rather that it is most difficult to  describe the frustration asso-
ciated with attempting to deal with: blatantly false statements pre-
sented as fact; completely inaccurate stories reported by the news
media as the truth; and accusations of unethical practices which
never occurred. The intensity of the feelings of those  in  opposition
can only be fully appreciated when one is on the receiving  end.
Words, written or spoken, are woefully inadequate as an explana-
tion.
  The fact remains that the majority of the  residents in the local
site area of any such project will be  unalterably opposed to  it.
They  have been educated very well by the news media  on the horror
stories of hazardous waste mismanagement. In many cases such as
this one, local residents are contacted early by organized, well-
financed opposition leaders and taught  how to delay,  and ultimate-
ly stifle,  a project. The presentation of the technical basis and other
facts about the project changed very few opinions.
                                                                                                 STATE PROGRAMS
                                                                                                                             519

-------
A BETTER WAY

  Ideally, the solution to the siting dilemma, the decision about
who should have a hazardous waste management facility near them
and why, lies in  the formulation  of fairly precise siting require-
ments by the regulatory agency. The "black and white" siting re-
quirements are the technical  criteria,  e.g.,  hydrogeology, eleva-
tion. The strict application of these technical requirements would
allow the applicant to follow a systematic  site selection  process
and locate a technically superior site, thereby reducing uncertainty
and economic risk. But  using the "black and  white" approach
(adhering strictly to technical siting criteria to grant  or  deny a
permit) would ignore the ever-present "gray area," public opinion.
  Local  residents' opposition cannot be ignored by the applicant
and will always be part of the hearing process. As mentioned, pub-
lic opposition often is not based on technical criteria. Therefore, in
order to insure that public opinion is considered and the  process
is equitable to all parties involved, the regulatory agency is forced
into the  position  of also formulating public participation criteria
along with the technical requirements.
  Now come the value judgments: How do you define an "affected
party"?  How much  weight should be given  to their arguments?
And on  and on.  Theoretically,  assuming these questions arc re-
solved, a process defining technical siting criteria and the boundaries
of public participation would be the best method for all parties
concerned; costs and frustration would be minimized if the parties
to a hearing knew the rules of the game going in.
                                                         Political Pressure
                                                           Nevertheless, under any arrangement, the decision making bur-
                                                         den, whether to permit or not to permit a facility, ultimately rests
                                                         with the regulatory agencies. On a practical note,  the following
                                                         holds true: regulatory agencies are funded by the state legislature.
                                                         Elected state officials very quickly hear from the irate and most
                                                         vocal opponents of a facility. They do not get hundreds of calls
                                                         and letters from individuals who support a hazardous waste facil-
                                                         ity, and the irate voters' position quickly becomes the position of
                                                         the elected official.
                                                           The state agency is then subjected to the hysterical  voices of
                                                         angry and misinformed area residents and  the unique persuasive
                                                         powers of legislators, whose re-election depends on  satisfying the
                                                         wishes and fears of area voters. Therefore, it is very  doubtful that
                                                         mandatory siting criteria  will be issued, or could even  be  work-
                                                         able, in Texas.  As  is known, other states have instituted this  mech-
                                                         anism and no regional treatment facilities have yet been permitted,
                                                         much less constructed. The need for proper hazardous waste man-
                                                         agement remains.
                                                         CONCLUSIONS
                                                           Any application for a  hazardous waste  facility,  regardless  of
                                                         location, will be intensely opposed. Application of well-defined
                                                         siting criteria for technical requirements and  public participation
                                                         could reduce the uncertainty of the applicant  and at least make a
                                                        dent in what is now a standoff.
 520
STATE PROGRAMS

-------
         BID  PROTESTS, CHANGE ORDERS AND  CLAIMS
              IN  THE  SUPERFUND  REMEDIAL PROGRAM
                                           LINDA GARCZYNSKI
                                      Hazardous Site Control Division
                                  U.S. Environmental Protection Agency
                                              Washington, D.C.
                                         LAURIE M. ZIEGENFUS
                                        Booz-Allen & Hamilton Inc.
                                             Bethesda, Maryland
INTRODUCTION

  This paper discusses bid protests, change orders and claims as
they  affect State-led  remedial  responses  conducted  under
CERCLA. Bid protests, change orders  and claims are procure-
ment-related issues of special concern during projects managed by
a State. They are, however, not unique to Superfund. They have
also been of concern to the wastewater treatment construction
grants program at the USEPA.
  The Superfund remedial program provides long-term, perma-
nent cleanup of the nation's abandoned hazardous waste sites.
Individual site cleanups  may be managed directly by  either the
USEPA or the affected State. If the State elects to manage the re-
medial response  at a site, it  enters into  a cooperative agreement
with the USEPA. The  cooperative agreement is an  assistance
vehicle which  transfers funds to the State and documents both
State and USEPA responsibilities for execution of the project.
  The recipient State agency then oversees the project, ensures that
it meets the guidelines laid out in the cooperative agreement and
manages the remedial funds  provided via the cooperative agree-
ment. At the same time, there is substantial Federal involvement
under a cooperative agreement  since the USEPA retains the ulti-
mate responsibility for ensuring that provisions of the agreement
are carried out.
  With the increasing number of remedial actions and the award of
more construction contracts by States, the number of bid protests,
change orders and claims may also be increasing. If not resolved in
a timely manner, these problems could seriously delay  the initia-
tion of remedial projects, affect project completion and lead to in-
creased costs.
  This paper discusses various avenues  for improving  the man-
agement processes  dealing with bid protests, change orders and
claims. The USEPA intends to  develop further detailed guidance
for use by Regional Offices and States patterned on the  Construc-
tion Grants Program guidance. This program provides funds to
municipalities for the construction of wastewater treatment plants.
Construction Grants is a fully delegated program in which the-
municipality directly manages the project, the State monitors per-
formance of the project and the USEPA provides and oversees the
State's implementation of regulations, policy and guidance. Much
of the following discussion is derived from the experience, regula-
tions and guidance developed for the Construction Grants Pro-
gram.

REGULATORY REQUIREMENTS

 A State must meet the requirements of several USEPA regula-
tions when implementing an  executed  cooperative agreement.
These regulations include the National Contingency Plan (40 CFR
Part 300), The General Regulation for Assistance Programs (40
CFR Part 30) and the Procurement Under Assistance Agreements
(40 CFR Part 33). Certain provisions of other regulations, as they
affect the above-referenced regulations, also apply (e.g., 40 CFR
Part 32, Debarment and Suspension).
  Procurement Under Assistance Agreements are USEPA's regula-
tions governing procurement of supplies, services and construction
by assistance recipients. Under these regulations for procurement
involving USEPA funds, a State may use its own procurement pol-
icies and  procedures if it certifies that its  system meets all the re-
quirements of this regulation. If the  State's procurement policies
and procedures do not meet all the requirements of 40 CFR Part
33, the State must use the requirements set forth in the regulations
and allow USEPA pre-award review of proposed procurement ac-
tions that will use USEPA funds (40 CFR Part 33 Appendix A).
  Under USEPA regulations, the State may procure contracts in
four ways: formal advertising, competitive negotiation, non-com-
petitive negotiation and small purchase procurement.
  For Superfund State-led remedial actions, formal advertising is
the preferred method  of procurement (40  CFR 33.910-Subpart
E). For remedial planning activities, a State may, at its option,
use the competitive negotiation method instead of formal advertis-
ing.
  Formal advertising requires, at a minimum:
•A complete, adequate and  realistic specification or purchase
 description of what is required
•Two or more responsible bidders who are willing and able to com-
 pete effectively for the recipient's business
•A procurement  that lends itself to the  award of a fixed-price
 contract
•The selection  of the successful bidder made principally on the
 basis of price (40 CFR 33.400(b)]
  Formal advertising or competitive negotiation for services max-
imizes  use of Hazardous Response Trust Fund monies  for sites
nationwide. Only in rare circumstances would the USEPA approve
the use of non-competitive negotiation.
  Even if recipients of CERCLA cooperative agreements  fully sat-
isfy the USEPA's procurement requirements, problems may arise.
The most common of these—and those that are currently of great-
est concern—are bid protests, change orders and claims.
  In reality, bid protests, change orders and claims are a series of
interrelated  issues associated with procurement. The Superfund
Program emphasizes the importance of preventing bid protests and
claims, to the extent possible, by the up-front preparation of qual-
ity solicitations and also stresses the need  to  effect meritorious
change orders through good project management.
                                                                                        STATE PROGRAMS
                                                      521

-------
BID PROTESTS

  Bid protests are written complaints filed by parties with a direct
financial interest affected by the State's procurement action. Since
complaints typically concern the solicitation (e.g., content or word-
ing of or deficiencies in specifications) or the actual award of con-
tracts, they are of concern to both the USEPA and the State. Bid
protests may significantly delay the initiation of a remedial project.
Thus, it is vital to avoid their filing when possible.
  A protestor may file an appeal with the USEPA after the State's
bid protest has been made and all available  administrative remedies
at the State level have been exhausted. The protestor may file an
appeal only when he feels his financial interest has been adversely
affected. The administrative process for the rapid resolution of bid
protest  appeals by  the USEPA are contained in  the  regulations
40 CFR 33.1105-1145. Limitations regarding timing and content of
the appeals are also defined.
  A State is required to develop procedures to promptly consider
bid protests (40 CFR 33.1110). Generally,  a State defers award of
the contract and delays initiation  of remedial activities pending  a
determination on the protest. Upon resolution of the  bid  protest
by the State, all affected parties have a limited period of time to
file a protest appeal with the USEPA.
  Upon receipt of a protest  appeal, the USEPA will  request the
State to defer award of the contract.  Although a  State is  not re-
quired to defer the award, it bears the risk that the costs of the con-
tract may not be allowable for Federal funding if the protest appeal
is upheld. If a protestor does  not agree to a request from the State
for a reasonable extension of the bid or bid bond period while the
protest or appeal is pending, the State or the USEPA can summar-
ily dismiss the protest or appeal.
  The USEPA's review of protest appeals  is limited to issues aris-
ing under the provisions of 40 CFR Part 33 and alleged violations
of State or local law or ordinances when  the USEPA has deter-
mined that there is  an overriding Federal requirement. A subcon-
tractor may only file a protest appeal  for certain issues related to
the award of that subcontract by a contractor (40 CFR 33.1115
and 33.295).
  The USEPA reviews the record  considered by the State and any
additional information regarding the basis of the appeal and ren-
ders a final decision. The USEPA's decision regarding a  protest
appeal is final  and  may not  be appealed  by  the protester or the
State. If a State does not comply with  the USEPA's  determina-
tion, the USEPA can take  action  against the  State under 40 CFR
Part 30 (40 CFR 33.1145).
  A State can minimize or avoid protests by developing high qual-
ity, unambiguous bid documents with a clear and accurate descrip-
tion of the technical requirements for any necessary materials, pro-
ducts or services. In developing specifications for the remedial pro-
jects, a State should incorporate a clear and accurate description of
the technical requirements for any necessary materials, products or
services and the required performance schedule.
  Through arrangements with the Corps of Engineers  (COE), the
USEPA can provide a State with  a mechanism to  do biddability/
constructability reviews at the State's  request. These reviews will
be conducted concurrently  with any internal review the State may
choose to conduct and will  not result in delays to the procurement
process. Because the COE performs remedial activities for Federal-
led remedial sites, their technical expertise  should prove extremely
useful to a State. The USEPA may also require as a condition of
the cooperative agreement,  the COE to perform such a review for
the State for projects which  are,  in the USEPA's opinion, tech-
nically complex or have severe time constraints.
  The USEPA  currently is exploring the possibility of  funding, as
a portion of cooperative agreements with  a State, the costs asso-
ciated with the State securing  independent services to perform bid-
dability/constructability reviews.
  Technical assistance provided by the COE or the use of services
of an independent reviewer could contribute to an effective bid pro-
tests prevention program implemented by any State. In the  case of
                                                        Superfund, expertise and experience is limited and bid protests and
                                                        subsequent protest appeals may hamper the timely implementation
                                                        of remedial activities.

                                                        CHANGE ORDERS
                                                          A change order is a written order issued by the State or its desig-
                                                        nated agent to its contractor authorizing an addition to,  deletion
                                                        from or revision of a contract, usually initiated at a contractor's re-
                                                        quest. Change orders are issued after execution of the contract. The
                                                        State  may also direct changes  in the contracts. Proper manage-
                                                        ment of change orders is a key element in avoiding increased costs
                                                        as well as contractor claims which are discussed in the next section
                                                        of this paper.
                                                          Administrative requirements  for management of change orders
                                                        including timing, form and provisions for contract adjustment ap-
                                                        pear in the model subagreement clauses of 40  CFR Part 33.1030.
                                                        These clauses delineate requirements for construction contracts
                                                        [clause 3(a)], contracts for services [clause 3(b)] and contracts for
                                                        supplies (clause 3(c)]. In addition, a clause exists to describe re-
                                                        quirements for changes resulting from differing site conditions
                                                        (clause 4).
                                                          Change order management practices which avoid significant in-
                                                        creases in contract  costs are essential to both  the State  and the
                                                        USEPA. States may choose to manage contracts directly or secure
                                                        the  services of a private construction oversight  firm. In the Super-
                                                        fund program, the design engineering firm frequently provides the
                                                        construction oversight services.  In addition, the USEPA will con-
                                                        sider funding services of a firm specializing in  change order man-
                                                        agement as part of the cooperative agreement with the State.
                                                          Change orders are often generated by the following conditions:
                                                        •Differing site conditions
                                                        •Errors and omissions in plans and specifications
                                                        •Changes instituted by regulatory agencies
                                                        •Design changes
                                                        •Overruns/underruns in quantities
                                                        •Changes in time of completion
                                                          In the course of remedial action, the  USEPA anticipates there
                                                        will be requests for change orders. Cooperative agreements with the
                                                        States routinely provide a contingency fund for the construction
                                                        contract for such change orders. This fund is often in the 5 to 15%
                                                        range. States are prohibited from using these funds, however, with-
                                                        out  obtaining USEPA concurrence.
                                                          Change orders may also be encountered during remedial plan-
                                                        ning phases of work at a particular site. States may rebudget ex-
                                                        isting contract funds in the cooperative agreement to pay for neces-
                                                        sary changes. However,  if change order costs exceed the funds in
                                                        the  existing cooperative agreement, the State  must request addi-
                                                        tional funds from the USEPA. State approval  of a change order
                                                        does not obligate the USEPA  to increase the amount (40 CFR
                                                        30.702) of a cooperative agreement.
                                                          Administrative and procedural requirements  for State manage-
                                                        ment of change orders are discussed in the model subagreement
                                                        clauses (40 CFR 33.1030). In addition, all negotiated change orders
                                                        exceeding $10,000 must have a State-conducted cost analysis.
                                                          If a change is deemed to be substantial, the cooperative agree-
                                                        ment  must be  formally amended [40  CFR 30.700(d)).  Change
                                                        orders  requiring an amendment to the cooperative agreement in-
                                                        clude:
                                                        •Significantly changed conditions at the site
                                                        •Changes substantially increasing or decreasing the funds needed to
                                                         complete the project
                                                        •Significant delay or acceleration of the project schedule
                                                        •Changes to the approved remedy
                                                          The  USEPA feels, however, that certain additional tools will
                                                        aid  a State in change order management. The USEPA will ask the
                                                        State to conduct an administrative and technical analysis of all in-
                                                        dividual change orders exceeding $100,000. This analysis should
                                                        consist of a review of the effects of such a change on the existing
                                                        scope of work.
522
STATE PROGRAMS

-------
  In the case of remedial action, the analysis must also take into
account the Record of Decision (ROD) selecting the most cost-
effective remedy. Should the change order cause the USEPA to re-
examine the selection of the remedy, the ROD may require amend-
ment prior to USEPA making funds available for the change order.
  Similarly, the USEPA will also require such an analysis from the
State when the aggregate costs of change orders exceed 10% of the
contract.
  In this way, USEPA will provide a State with the opportunity to
assess the effect of change  orders on the remedial activities. The
State must promptly consider all proposed change orders and issue
those that are meritorious. This process includes an investigation of
contractor reports of differing site cdnditions to see if they will re-
quire a change  order.  These activities are within the scope of the
contract management tasks  and are included as a standard task
in most cooperative agreements.
 CLAIMS

  Claims consist of requests by the State's contractor for changes
 (e.g., additional time and/or costs) which have been originally re-
 jected by the State. These claims, obviously are of significant con-
 cern to both the USEPA and the State.
  The State is responsible for the satisfactory completion of the
 contract for the work  outlined in the cooperative agreement (40
 CFR 33.210). However, the  State may still  encounter the issue of
 claims when the project has  been conducted in a thoroughly satis-
 factory manner.
  The development of a claims policy and the procedures for ad-
 ministering this policy  are of vital concern to the  Superfund pro-
 gram. The costs of claims are subject to budgeting considerations
 and must be analyzed  in light  of funding priorities for all Super-
 fund activities.
  Claims may be encountered by a State %)Oth in contracts for serv-
 ices and for construction. Many claims that States have encount-
 ered result from:
 •Defects in plans or specifications
 •Differing site conditions
 •Inadequate construction inspection and management
 •Failure to promptly and fairly address contractor grievances, re-
  quests for time extensions or other problems
 •Failure to  enforce contract provisions on scheduling and com-
  pletion time
 •Failure to  negotiate time extensions and/or delay costs, if any,
  associated with change orders
 •Failure to mitigate effects of delay
 •Unusually severe weather conditions
 •Strikes
 •Acts of God
  Consequently, good project management practices have  been
 identified which could reduce the number of change orders and the
 occurrence of claims for any type of contract in the State-led re-
 medial program. Some elements of good management include:
 •Recognizing the importance of scheduling as  a key management
  tool by specifying that the contractor provide a realistic and ade-
  quate schedule commensurate with the complexity of the project.
  Enforcing the schedule provisions and requiring periodic updating
  to show the adjusted project progress and completion date are
  equally important.
 •Maintaining a  full and  completely documented record of all
  aspects of the work (such as photographs) and a daily log of work
  progress, personnel and equipment on site.
 •Demonstrating a  knowledge and. understanding of  common
  sources of disputes or situations likely to result in claims during
  construction and exercising effective techniques to  avoid  such
  situations.
 •Exercising  effective .management of change orders, resolving all
  costs and any necessary contract time extensions associated with
  each change order as the change order is executed.
•Providing timely responses to contractor requests for direction,
 clarification and adjustment.
•Instituting measures to ensure accurate and complete plans and
 specifications  (e.g.,  biddability/constructability reviews) and
 holding pre-bid conferences.
  The USEPA can, by amending a State's cooperative agreement,
fund a portion of the costs that  a State incurs in analyzing the
merits of claims as well as the costs associated with negotiating a
settlement or defending itself against  claims. However, a State
must request such claims management  funding  from the USEPA
in advance of resolving the claims. The USEPA then reviews the
schedule, budget and scope of work required for claims manage-
ment by the State and assesses whether the claims resulted from
poor management or other factors. This determination is the basis
for the USEPA's decision to fund claims management costs.
  After the State receives the award of the cooperative agreement
amendment, the process of claims management or defense against
claims may proceed. The following suggested actions will aid the
State in resolving claims:
•Take immediate steps to mitigate further costs  being incurred by
 the contractor, or any other party, due to the claims issue
•Perform  a timely,  complete and thorough review  of the  issues
 raised by the claims to determine the degree of merit that each
 issue may have
•Negotiate with the contractor on the issues in a good faith attempt
 to resolve each issue
•Make a renewed effort to negotiate a  fair  and  reasonable settle-
 ment of the meritorious issues and a reduction or elimination of
 the issues found to be without merit
•Maintain a fully and completely documented record of the  nego-
 tiation process used to resolve the claim
•Provide a high degree of attention to dispute resolution [40 CFR
 33.1030 (clause 7)]
  During the claim(s) resolution process, the State may choose to
consult with the USEPA. Because the State is responsible for all
meritorious contractor claims, the USEPA must  carefully evaluate
the extent of the USEPA's interest in awarding the State the costs
of claims.
  The State may choose to submit a cooperative agreement amend-
ment request to the USEPA before a final settlement takes place.
This will enable the State to determine the USEPA's evaluation
of the extent to which the USEPA may award costs allocable and
allowable to the project prior to reaching a settlement with the con-
tractor.  Should the State negotiate a settlement,  the USEPA must
determine whether the costs associated with the  claim are (1) elig-
ible, allocable and allowable (40 CFR 30.200); (2) within the  scope
of work agreed upon; (3)  consistent with the record of decision;
or (4) the result of differing on-site conditions which caused immi-
nent and substantial endangerment requiring immediate attention.
  Claims that alter the cost-effectiveness analysis and selection of
the remedy may require a supplemental ROD. Approval of such
a ROD must precede award of funds for a cooperative agreement
amendment.
  When funds become available, the USEPA may choose to award
to the State, via an  amendment to the cooperative agreement, the
costs of the claim(s) determined to be allowable. Thus, the USEPA
has made claims prevention, claims management and claims reso-
lution an  essential element of managing  the costs of cleanup at
Superfund sites.
CONCLUSIONS
  The effective management of procurement and procurement re-
lated issues is a major segment of the State-led remedial program.
Issues resulting from bid protests, change orders and claims have a
direct effect on the USEPA, the State and the contractor.
  A State can minimize or avoid protests with high quality, unam-
biguous bid documents that contain a clear and accurate descrip-
tion of the technical requirements for any necessary materials,
                                                                                                STATE PROGRAMS
                                                          523

-------
products or services. To ensure that Superfund remedial project
requirements are adequately explained and schedules clearly delin-
eated, the USEPA,  via an interagency agreement,  is offering the
assistance of the U.S. Army Corps of Engineers to perform bid-
dability/constructability reviews for a State. The USEPA is also
considering providing funds to a State to secure independent third
party services  to  conduct  biddability/constructability reviews.
In certain instances, because of technical complexity or scheduling
concerns, the USEPA may require a State to submit bid documents
forCOE review.
  With  regard to change orders, the USEPA emphasizes the  im-
portance of change  order management.  Good change order man-
agement will provide a mechanism to control costs of a project and
a mechanism to avoid the filing of claims.
  Major checkpoints in the process will include: oblaining USEPA
approval to draw down construction contingency funds; technical
review and concurrence on the State's analyses of change orders
in excess of $100,000; and notification and provision of an analysis
to the USEPA of the effect of change orders when change orders
exceed 10% of the contractual  funds. These checkpoints provide
the USEPA with the ability to monitor project progress and costs
and provide the States with funds and management tools to effec-
tively pursue the objectives of the remedial project.
  Claims remain  another area of concern to  the Superfund pro-
gram. The USEPA proposes to provide funding and technical sup-
port to a State to prevent, manage and resolve claims via the co-
operative agreement process. As mentioned previously,  effective
change order management is a vital element in the prevention of
claims. By funding a State's claims management costs and provid-
ing Agency expertise, the USEPA hopes to furnish a State with an
enhanced capacity to resolve claims and to defend against claims.
  The USEPA will  consider funding claims allowable to  the pro-
ject in an amendment to the existing cooperative agreement. In this
way, a State will be recompensed for expenditures of time and
effort in good project management.
  The  USEPA will provide  more detailed information to States
via  an addendum to the  Slate Participation in  the  Superfund
Remedial Program  manual in the near future. The  USEPA will
also consider additional methods for improving the procurement
process and change order management and for  minimizing bid pro-
tests, protest appeals and claims. The USEPA will also determine
whether regulatory changes or additions are necessary based on the
effects of implementing the above-detailed initiatives.
  The  USEPA hopes  implementation  of these  measures  will
streamline projects, prevent some cost increases and allow  projects
to proceed to completion within the scheduled time frame.
524
         STATE PROGRAMS

-------
   S-AREA:  NEGOTIATED  REMEDIAL  PROGRAM FOR THE
            NIAGARA FRONTIER'S MOST  COMPLEX  SITE

                                           C. KENNA AMOS, JR.
                            U.S.  Environmental Protection Agency,  Region II
                                             New York, New York
                                          MURRAY E. SHARKEY
                      New York State Department of Environmental Conservation
                                              Albany, New York
INTRODUCTION

  In 1978, a chemical sludge was discovered in the intake struc-
tures of the Niagara Falls, New York, drinking water treatment
plant. Further investigation revealed that the source of these con-
taminants was an adjacent inactive landfill, known as S-Area (Fig.
1). Owned by the Occidental Chemical Corporation, formerly the
Hooker Chemicals and Plastics Corporation, the site was also con-
taminating the Niagara River, an international body of water. In
this paper, the authors discuss the history of the case and present
the most important components of a proposed settlement  agree-
ment that has been submitted to the Court.

CASE HISTORY
  To  remedy the problems created by S-Area, the United  States
Department of Justice, acting on behalf of the USEPA, filed a law-
suit against Hooker on Dec. 20,  1979. This civil action sought in-
junctive relief from the imminent and  substantial endangerment
that arose from the contaminants that had escaped and were  escap-
ing from the site.1'2 This action, one of four taken against Hooker
in Niagara Falls, was one of the first in the country regarding a haz-
ardous waste site. The State of New York (State) joined the  action
and became a co-plaintiff on Nov. 18, 1980. Earlier that same year,
the City of Niagara was named as a co-defendant.
  On Jan.  10, 1984, after more than three years of negotiations,
the parties  filed a proposed stipulation and judgment,3 to resolve
the litigation, with the U.S. District Court for the Western District
of New York. Thousands of hours of work went into the negotia-
tions; the USEPA spent more than $2,000,000 in consultant fees.4
More than 100 negotiating sessions were held.
  Several groups sought to intervene in the case.  Although local
environmental groups, with  both American  and Canadian  mem-
bers, were denied such status, the Court did grant intervenor status
to the Canadian Province of Ontario. (The environmental groups
have appealed the Court's decision.) A diplomatic note was sent to
the United  States by Canada to express  its criticisms of the  agree-
ment. Technical meetings were held with representatives of both
the Canadian federal and provincial governments. Through the
Niagara Frontier Agenda, a bilateral consultative group, frequent
discussions of the case were held with the Canadians.
  An evidentiary hearing (similar to a trial was held in Court from
Apr. 30 to May 3, 1984. At this hearing, the terms of the agreement
were presented to the Court and Ontario's arguments were heard.
As of Sept.  1, 1984, the Court's decision was still pending. Pur-
suant to the agreement, official implementation activities would be-
gin after a required 60-day appeal period which would follow a de-
cision approving the agreement.
 SITE HISTORY
  The site was used by Hooker as a dump from 1947 until 1961.5
 Built on land that was partially reclaimed from the Niagara River,
 the site lies at the southeast corner  of Hooker's Buffalo Avenue
 plant (Figs. 1 and 2). On it are situated two settling lagoons util-
 ized in a nearby process.  Eastward,  directly across 53rd Street, is
 the Niagara Falls drinking water treatment plant: a 64,000,000 gal/
 day facility with water treatment units known as plant A (ca. 1953)
 and plant B (ca, 1911).
  Beginning in 1947, Hooker dumped approximately 63,100 tons
 of inorganic and organic chemicals at S-Area (see Table I);6'7 the
 predominant compounds were chlorinated hydrocarbons.  These
 solid or  liquid chemicals were deposited in barrels, in other con-
 tainers (including at least one railroad tank car) or in bulk. Other
 deposited materials included construction wastes.
 SETTLEMENT AGREEMENT OVERVIEW
   The agreement presents a program to remedy the problems. In-
 cluded are elements of both a remedial investigation and a remed-
 ial action program. While general guidance is given, specific engi-
 neering  details will be  developed only after more extensive field
 data are gathered. Thus, the agreement is dynamic.
   The settlement agreement outlines activities  that may require 4
 to 8 years or more for complete construction. Thereafter, mainte-
 nance and monitoring  activities will continue  until  the 35th and
 38th anniversaries, respectively, of  the Court's approval of the
 agreement.  If the site continues to endanger the public health and
 welfare and the environment,  the terms of the agreement will be
 extended until the endangerment exists no longer. Thus, the agree-
 ment is highly protective.
  The settlement agreement was patterned on the work that the
 Federal and State governments performed in another Hooker case
 in Niagara Falls—the Hyde Park landfill. The additional  exper-
 ience at  S-Area and its settlement agreement document  have al-
 ready been used  by USEPA and the State in negotiations with
 Hooker  and the  Olin Corporation  in the 102nd Street case in
 Niagara Falls. Thus, the document is useful elsewhere.
  Finally,  the settlement agreement  is  part of a comprehensive
 remedial effort at Hooker's Buffalo Avenue plant. The State has
 initiated a lawsuit against Hooker regarding multiple hazardous
 waste  sites there [State of New York & Henry Williams v. Occi-
dental Chemicals Corp., Civ. Act No. 83-1393 (W.D.N.Y., Dec. 8,
 1983)]. All parties recognize  that remedies  from  these  actions
should be designed and operated compatibly, if possible.8 Thus, the
document both affects and is affected by other actions.
 Note: The views expressed herein are those of the authors and do not necessarily represent those of
 the USEPA of the State of New York.
                                                                                         STATE PROGRAMS
                                                                                                                  525

-------
     LtStMB
              J HOOKER'S BUFFALO  AVENUE

              -EXmtW  S-AREA FENCELINE
     —	MMKINO mil* TREATMENT  PLANT  •OVNOART

     ............. MflEAKWALL
                                                                Figure 1
                                               Geographical Location of S-Area Landfill Site
                                                                          Wulr
                                                                          Category
                                                                                                   T»bkl
                                                                                          S-Area Chemical Inventory
                                                                                     Phydcal
                                                                                     Suit
Efttnuled Quality
(Tons)
                                                                       Organic Prosperous
                                                                         Compounds             U S

                                                                       Miscellaneous Acid
                                                                         Chlorides             US

                                                                       Phenol Tars              L
                                                                         (Including chloro-
                                                                          benzenes)
                                                                       Ihionyl Chloride         L

                                                                       HET Acid
                                                                         (Chlorendic Acid)      L,S

                                                                       Miscellaneous Chlor-
                                                                                                         200


                                                                                                         400

                                                                                                         800



                                                                                                       4,100


                                                                                                         500
             PftO»OSCD SITC t«NKIIK WALL

             OUSTIN4  S-AftCA FCNCCLfie

             OKINKINQ WATIK TKIATMCXT PLANT lOUNOAftY

      until  WIIAXWALL


                            Figure 2
                 S-Area/Water Treatment Plant Site
                                                                                                                    ConUlKr
                                                                                                                       D.B
                       D,B
                                                                                                                       D.B
1 nations
Oodecyl Me reap tans
Tr ichlorophenol
Bentoyl Chloride
Liquid Dlsul fides/
chlorotoluenes
Metal Chlorides
Hexachlorocyclopenta-
diene
Chlorobenzenes
Benzyl Chlorides
Thiodan (Bndoaulfan)
Sulfldea
Miscellaneous (10% of
above)
L,S
L,S
L,S
US
us
S
us
us
us
us
s
TOTAL
400
600
200
3,300
2,200
900
17,400
1 9, 900
1,600
700
4,200
5,700
63,100
0,8
D
D
D.B
D,B
D
0,8
D,B
D.B
D,B
0

                                                                            Legend)  L - Liquid,  S - Solid, D • Drumed, B • Bulk
526
STATE PROGRAMS

-------
IMPORTANT/INNOVATIVE FEATURES
OF THE SETTLEMENT AGREEMENT

  First, the proposed agreement' is flexible. Any specified remedies
and monitoring systems can be modified, based upon the evalua-
tion of actual field data. The agreement requires adequate data
before any final decisions  can be  made regarding any remedial
measure.
  Contained in the agreement is a unique mechanism which com-
bines flexibility with the requirement for adequate data: requisite
medical technology  (RRT).  Through a phased approach, Hooker
will gather and evaluate data to determine more fully the  nature
and extent of off-site contamination. If those data are inadequate,
Hooker must gather more data. More importantly, Hooker must
initiate an RRT study to find a remedy for an identified problem.
  The RRT considers human endangerment as well as the prac-
ticability of implementation. Hooker is required to  consider all
available information  in assessing  off-site contamination as well
as any remedy. Thus, the RRT process uses a fully-developed
scientific data base to  find and to designate appropriate solutions,
if any. Prototypical remedies can be installed to  evaluate  such
remedies by gathering more data.
  Computer models will be used innovatively through  the agree-
ment. Four groundwater flow models were used during the nego-
tiations to evaluate the impact of the  proposed remedies. The
USEPA modified the USGS (Finder) model and then verified those
predictions with another model that was developed. Such model-
ling will be used, in the program, to  insure the effectiveness of these
remedies.
  Another way in which computer models were used, and  will be
used,  regards the innovative upward hydraulic gradient concept.
A two-phase transport model,10 suggested by the USEPA (Finder)
and developed by Hooker, will be used to verify that the required
upward gradient is being obtained.
  Another feature is the extensive landfill monitoring program that
will be required. Its components—hydraulic,  chemical and tracer
systems—will  provide redundant protection by monitoring the
effectiveness of the remedies.
  The environmental  health and safety  plan" is another impor-
tant feature. It will protect on-site workers as well as off-site work-
ers (Hooker's and others) and nearby residents.
REMEDIAL PROGRAM ELEMENTS
  To  remedy the problems caused by S-Area and to protect the
users of the City's drinking water supply, the proposed  settlement
agreement contains five main activities:12
•Containment and, to the maximum extent possible, collection of
  the deposited hazardous materials
•Employing the RRT concept that will be used to develop remedies
  for chemicals that have migrated off-site
•Correction of damages to and cleaning of the Niagara Falls drink-
  ing water treatment plant, including the property
•Long-term monitoring of the landfill and treatment plant contain-
  ment and collection systems and the finished drinking water qual-
  ity
•Long-term maintenance of all remedies
  Another critical component of the agreement is a fiscal guarantee
from Hooker that the  necessary funds will be available to meet its
obligations to fulfill the terms of the agreement.13 That guarantee,
the previously-mentioned environmental health and safety plan and
the five main activities comprise the terms of the proposed agree-
ment.
  The following discussion of the  major  program  elements is
divided into the two functional elements:  the S-Area landfill, itself,
and the Niagara Falls drinking water treatment plant.

THE LANDFILL CONTAINMENT PROGRAM
  According to the settlement agreement, the purpose of the con-
tainment program14 is "...to identify and where required, using
requisite remedial technology, to contain or collect chemicals which
have migrated into soils, bedrock, sediment, surface runoff waters,
groundwater, and air from the landfill site, when detected at levels
at or above specific survey threshholds."15

Site Geology and Hydrogeology

  The site is composed  generally  of geologic materials  that are
categorized as the overburden and the bedrock.16 Overlying the
bedrock, the overburden consists  of two layers: the uppermost
unconsolidated layer composed of sand, silt,  clay and fill (e.g.,
cinders, stone, slag, wood, dirt, etc.); and the lowermost glacially-
derived clay/till layer. The uppermost layer's thickness is approx-
imately 30 to  35  ft; the  lowermost's  ranges from approximately
0 to 20 ft.
  The significant bedrock units include the  uppermost Lockport
dolomite formation and the underlying Rochester shale formation.
The topmost 15 ft of the  Lockport formation are highly fractured
and serve as a pathway for chemical migration. The bedding planes
in both formations  slope gently (30 ft/mile)  to the south, toward
Canada.
  The upper zones of the  overburden and the bedrock are the most
permeable. There is at least one discontinuity (or hole) in the clay/
till layer through which chemicals have migrated into the bedrock.
  Two  major  lateral flow zones comprise the groundwater  flow
zones. The flow in  the overburden is generally from  the north to
the south, toward the Niagara River (Fig. 3). In the bedrock, the
flow is generally in the opposite direction (Fig. 4).
Field Investigations

  The proposed settlement agreement requires initial surveys and
studies.17-  18 According to a predetermined logic,  wells would be
drilled in the overburden, the bedrock and the utility beddings in
and around the known disposal area.  Sampling and analyses will
then be made  for both general and specific indicator chemicals.
The resulting data will be analyzed to assess  the spatial migration
of chemicals in the geological zones.
          LEGEND

          EXISTING S-AREA FENCELINE
         1 DRINKING WATER TREATMENT PLANT BOUNDARY
          FLOW DIRECTION
                           Figure 3
Generalized Flow Directions in the Overburden Water Table at the S-Area/
                   Water Treatment Plant Site
                                                                                                STATE PROGRAMS
                                                                                                                           527

-------
                                                                                             Table 2
                                                                       Average Composition of S-Area Non-Aqueous Phase Liquid
                           Figure 4
    Generalized Flow Directions in the Bedrock Aquifer at the S-Area/
                   Water Treatment Plant Site
  These surveys and studies divide the contamination into the two
distinct phases, aqueous and non-aqueous, discovered in previous
investigations. The  aqueous phase comprises those soluble chem-
icals which  flow with the groundwater. The  non-aqueous phase
liquid ("NAPL") is that mixture of relatively insoluble chemicals
which is more dense than water and which can move independently
of the hydrogeologic conditions" (Table 2).
  Concurrent with chemical sampling, piezometric elevations will
be determined, and the geological and hydrogeological characteris-
tics of soil and rock will be examined. In particular, special con-
sideration will be given to the geological characteristics of the in-
dustrial intake-pipe trench.  This trench,  which  contains large-
diameter pipes (up to 72 in.) that transmit river water to the chem-
ical plants near the river, crosses the S-Area boundary at two loce •
lions (Fig. 5).
  This survey and study program will describe the three classes of
parameters influencing remedial activities at this or any other inac-
tive hazardous waste site:
•The levels of chemistry present
•The pathways for migration
•The dynamic influences
  Full description of these elements will permit both the considera-
tion and the design of specific remedies.

Selected Remedial Concepts

  The following general remedies were considered: excavation and
disposal; excavation, destruction and disposal; in situ destruction;
or various containment schemes.
  The first three concepts were examined extensively during nego-
tiations. As an  initial  remedy at this site, excavation was deemed
infeasible for several reasons. The mere quantity (up to 10' bank
yd') of soil and bedrock requires immense resources for either final
disposition (including sites),  thermal  destruction capabilities or
both.  Large volumes of contaminated groundwater,  requiring
treatment and disposal, will be generated. The risk to both work-
ers and the community-at-large would be unacceptable.
  In situ  destruction  (e.g., incineration, chemical  or  biological
techniques, etc.) was not commercially available to accommodate
this type and quantity of waste. Thus, it was decided to develop a
Average
Conpound (% of
Tetrachlorobenzenes*
Tr IcKlorobenzenes*
Tetrachloroethylene
Hexachlorocyclopentad lene
Octachlorocyclopentene4
Pentachlorobenzene
Carbon Tetrachloride
Toluene
Hexachlorobutadiene*
Honochlorobenzene*
Hexachlorobenzene'
Monochlorotoluenes
Oichlorobenzenes
Hexachloroethane
Trichloroethylene
Compoaition
Total %)
37.2
15.5
7.5
6.5
10.0
6.4
1.5
1.8
3.6
3.7
2.8
1.2
0.9
0.5
0.3
Water Solubility
(milligran/literi
0.4
30.0
200.0
2.0
<0.1
<1.0
800.0
500.0
2.0
500.0
0.02
20.0
123.0
50.0
1,100.0
                                                                 TOTAL PERCENT
                                                                                             99.4
                                                        I S-Area Indicators           72.8

                                                        Legend: 'Chemicals that are used as S-Area indicators. Data taken from Reference 20.
                                                        ^Va-.gsp^^-t :r-'-~.
                                                            /Łf	-"^arvN
                                                                                   Figure 5
                                                                        Location of Barrier Walls and Plugs
528
STATE PROGRAMS

-------
containment system composed of a barrier-wall (with hydraulic
controls), a clay (bottom) confining layer and surficial capping.
  Envelopment or encapsulation of contamination at inactive haz-
ardous waste sites is fairly common. At S-Area, however, in order
to develop a practicable and reliable containment system, select-
ing this remedy required rigorous planning, modelling and evalua-
tion. Two site-specific concerns  adding to the complex nature of
this landfill are the presence of NAPL and the absence of a con-
tinuous, low-permeability, confining stratum. These two concerns
were addressed by using another site-specific hydrogeologic con-
dition advantageously: the bedrock aquifer, which is influenced by
the adjacent Niagara River, will be used as a barrier to the down-
ward  flow of chemicals from the site (Fig. 6).
Containment System Components

  The containment system" will consist of the following:
•Barrier walls
•Drain-tile collection system
•Barrier plugs
•Surficial capping
•Clay (bottom) confining layer
•Upward (hydraulic) gradient
  These are fully described in the proposed settlement agreement.
  The agreement provides for a 2-ft barrier wall that circumscribes
the site.  This wall, having a maximum permeability of lxlO~6 cm/
sec, extends from the land surface down to a depth of 1 ft into the
clay layer. While this barrier will impede the lateral flow of chem-
icals from the  site, its primary purpose will be to significantly re-
tard, and thus reduce, the flow of groundwater into the contain-
ment system. This flow reduction is a major consideration in the
overall containment scheme.
  A drain-tile collection system will be used with the barrier to
"maximize the containment (and collection) of aqueous and non-
aqueous phase  liquids located within  the site barrier wall."23
The collected liquids will be adequately treated and subsequently
discharged.
  Where the industrial intake-pipe trench crosses the site bounda-
ries (Fig. 5), plugs will be emplaced to reduce the trench's ability to
act  as  a pathway  for  chemical1  migration.  Low permeability
(lxlO~6 cm/sec) grout will be used. These plugs will extend down-
ward to a depth below the bedding material in the trench.
  Complete  surficial capping of sites is another standard tech-
nique modified  for use  at S-Area. There  are two actively-used
process-waste settling lagoons situated on  S-Area (Fig. 5); their
continued operation  atop  the contained landfill  will  not be ex-
cluded.  To eliminate their influence upon  the landfill, these la-
goons, if operated, will be reconstructed with a protective system
consisting of a  synthetic liner,  a coarse-sand layer, a  drain-tile
collection system and a clay liner. Thus, even if the lagoons are
operated, there  will be the equivalent of complete surficial cap-
ping. Once installed, the cap will greatly retard the influence of sur-
ficial water on the landfill.
  Basic to most containment systems is  a barrier  which prohibits
downward flow  of chemicals. Geologic conditions usually provide
such a barrier.  As mentioned previously, a low-permeability clay
layer underlies most of S-Area; in at least one location, though, this
stratum is absent. Required testing will define this area of discon-
tinuity as well as the integrity of the existing clay (bottom) confin-
ing layer. Where the clay exists in sufficient depth, its imperme-
ability will act to retard chemical flow.  Total downward vertical
migration through the discontinuity will be prevented, nonetheless,
as will be seen.
                       "^ If ij'in™ — — —
                        \      •
                        •     '»   I
                                             Gwi
                                            clay & till
      H

      After
                 clay cap
                                                                                                           clay * till
                                                                                                            lockport
                                                                         *'tar        c|a, cap-
                          Figure 6
Impacts of Drains in Causing Inward and Upward Groundwater Flow.21
(The water levels are lower in wells inside the landfill than in those wells
       outside of the landfill, and below the clay/till layer.)
                           Figure 7
           Concepts of the Upward Hydraulic Gradient."
 (The water level difference between wells in the overburden and bedrock
        must be large enough to cause upward flow of NAPL.)
                                                                                                 STATE PROGRAMS       529

-------
  Another hydrogeological barrier used to contain chemicals is
the difference in hydraulic pressure between aquifers. In S-Area,
at or  near the discontinuity, an  upward gradient  will be created
through the drawdown  of the drain-tile collection system (Figs.
6 and 7).  Installation of sufficient drain tiles will lower the over-
burden's groundwater elevations. This action will cause the upward
pressure exerted by the  bedrock aquifer to become greater than
that pressure in  the overburden. Thus, the contaminants will be
either suspended or moved upward toward the drain tiles (the more
likely occurrence). While the chemicals will be collected, the pur-
pose of the upward gradient will not be to flush or eventually re-
move all chemicals from the site; its purpose will be to reverse  the
flow of groundwater, back into the site, thus preventing the down-
ward migration of chemicals.
  The upward hydraulic gradient concept is essential to the overall
containment  of both aqueous and non-aqueous phase chemicals.
As mentioned earlier, the feasibility of this upward gradient has
been demonstrated through both transport and flow modelling per-
formed by Hooker,  by the USEPA and by the Slate. Two of these
models will be used to verify the performance of this innovative
remedial concept.
Off-Site Remedial Concepts
   In addition to the containment of chemicals at the site, the pro-
posed agreement provides remedies regarding three other concerns
at or  near S-Area: an adjacent disposal site just north of S-Area
(the Northern site); contamination that has migrated  off-site in
the overburden, and contamination that has migrated off-site in  the
bedrock.  A  Northern Containment System similar to S-Area's
would be constructed." Even though the thickness of the clay layer
is expected to suffice for vertical containment, this assumption will
be verified by testing.
   off-site migration into the bedrock and into various strata out-
side the S-Area's bounds will be studied  to develop remedies.1'
These studies, comprising a program similar to  a  remedial inves-
tigation, will determine an RRT. If a remedy is  warranted, based
upon the  findings of the studies,  either the parties will agree to its
terms or it will be subject to a Court-determination. Thus, ancillary
issues can be resolved without delaying installation of major com-
ponents of the site containment system. Such containment neces-
sarily precedes some major remedial activities at the water treat-
ment plant.
Implementation

  The containment system will be a part of the remedy at S-Area.
Installation of that  system will be predicated  on specific designs
based upon gathered field data. The plans, specifications and pro-
tocols will be submitted for USEPA and State approval prior to
any construction. After such approval, the governments will over-
see the installation of the remedies. At the end of 38 years, if an en-
dangerment remains, these systems will continue to operate until
that endangerment ceases.
Maintenance and Monitoring

  Hooker must maintain the installed remedies for the term of the
agreement."  In addition to hydraulic  monitoring, both chemical
and tracer monitoring systems will be  required.  After conducting
field tests, Hooker will design these systems. Guidance has been
given  regarding the number of wells and the approximate locations
for the  final  systems. If, however, Hooker demonstrates the  in-
feasibility  of  such monitoring, it  will not be required: alternatives
will be required. If  the systems are installed, many years of data
will be reviewed to  determine whether the systems are providing
the required type and quality of data. All three systems monitor
the long-term effectiveness of the containment systems.11

DRINKING WATER TREATMENT PLANT
REMEDIAL PROGRAM"

  The remedies will be designed to protect the public water supply
against  past or  potential  future  contamination from S-Area.
                                                       Hooker will install or undertake them "without altering existing
                                                       structures, and without adversely affecting the City's ability to ade-
                                                       quately process  and supply  finished water for  distribution."111
                                                       These  remedies  are divided generally  into three major  com-
                                                       ponents:"
                                                       •Measures that eliminate the potential for chemicals to infiltrate
                                                        the treatment plant structures and piping
                                                       •Measures that prevent contamination from  entering the bedrock
                                                        intake system through its floors or walls
                                                       •Measures that remove past contamination from the  treatment
                                                        plant structures and piping
                                                         These measures  will  be augmented by both  maintenance and
                                                       monitoring programs.27- 32 There will be a separate maintenance
                                                       agreement between Hooker and the City. Hooker will monitor the
                                                       collection systems to ensure both free-flow conditions and proper
                                                       operation. Groundwater elevations around the water-bearing struc-
                                                       tures will also be monitored.  Finally, the plan requires  stringent
                                                       water-quality monitoring of both the intake-system water, to dis-
                                                       tinguish S-Area's contribution from  the river's loading, and the
                                                       finished (distribution) water.  Any detected problems or elevated
                                                       chemical levels (above an established background level) will prompt
                                                       immediate corrective action.
                                                         A highly-impervious surficial cap will be placed on the contam-
                                                       inated portions of the property.


                                                       CONCLUSIONS
                                                         Three years of negotiations  have produced  an agreement setting
                                                       forth a workable solution for the  problems created by S-Area.
                                                       This negotiation process represents a practical resolution of this
                                                       case. Even though  the  years of negotiations  might seem  too
                                                       lengthy, there was  no guarantee that  a more satisfying solution
                                                       could have been obtained by taking the case of trial or by invoking
                                                       the  statutes of CLRCLA. The experience gained at S-Area and the
                                                       proposed agreement produced in the case have already been used
                                                       by the USEPA and the State in other negotiations.
                                                         The USEPA invested significant extramural funds, although the
                                                       burden of supplying the data lay with Hooker. And, even though
                                                       the  USEPA's and the State's oversight will be costly, the remedial
                                                       program  will cost  Hooker a  minimum  of  approximately
                                                       $36,000,000. An environmental hazard of national significance on
                                                       the  Niagara Frontier will be  remedied. The settlement agreement
                                                       represents, therefore, a milestone in joint efforts of the USEPA
                                                       and the State.
                                                         Thorough  investigation and rigorous evaluation of all elements
                                                       of concern allowed the proposal of an overall program  (strategy)
                                                       to protect the public health and welfare and the environment. The
                                                       program will continue  as  long  as an endangerment exists. The
                                                       remedies set forth in the settlement agreement are dynamic, useful
                                                       and effective. With regard to the hazardous wastes, this program is
                                                       designed  to  achieve  total containment  and maximum feasible
                                                       collection of chemicals.

                                                       ACKNOWLEDGEMENTS
                                                         Many engineers and scientists were involved  in this case.  The
                                                       authors extend thanks to their governmental colleagues with whom
                                                       they worked on this case and, in particular,  to the following con-
                                                       sultants: Drs. Samuel Fogel and Neil  Shifrin, Cambridge Analyti-
                                                       cal  Associates,  Boston,  MA; Dr. Benjamin Mason, ETHURA,
                                                       Inc., Derwood, MD; Dr. Charles Faust, GeoTrans, Inc., Reston,
                                                       VA; Drs. George Pinder and Eric Wood, Princeton University;
                                                       Dr. Robert  Harris, ENVIRON, Inc., Washington, DC; Walter
                                                       Morris, Stephen Talian and Albert Gray, Gannett Fleming Water
                                                       Resources Engineers, Inc., Camp Hill, PA;  Dr. Brenda Kimble,
                                                       Amphion Associates, Inc., Alexandria, VA; Dr. Denys Reades,
                                                       Colder Associates, Inc., Mississagua, Ontario,  Canada; and the
                                                       Hooker consultants, including the late Dr. Philip  Levins of A.D.
                                                       Little, Inc., Cambridge, MA.
                                                         The authors thank and commend the legal staffs of all parties.
 530
STATE PROGRAMS

-------
REFERENCES

 1. Memorandum  Concerning Proposed  Stipulation  and Judgment
   Approving Settlement Agreement (re: United States v. Hooker Chem-
   icals & Plastics Corp. et al,  Civil Action No.  79-988, W.D.N. Y.,
   December 20,1979), Jan. 27, 1984,1.
 2. Complaint for Injunctive Relief and Restitution (re: United States
   v. Hooker Chemicals &  Plastics Corp., et al., Civil Action No. 79-
   988, W.D.N. Y., Dec. 20,  1979), Dec. 20, 1979, 1.
 3. Stipulation and  Judgment Approving Settlement  Agreement  (re:
   United States, the State of New York v. Hooker Chemicals & Plastics
   Corp., et al., Civil Action No.  79-988, W.D.N. Y.,  Dec. 20, 1979),
   Dec. 3,1983.
 4. Schafer,  Jacqueline A., Testimony before the Hon. John T. Curtin,
   Senior Judge,  U.S. District Court for Western New York: Hearing
   Proceedings (re: United States, et al.  v. Hooker Chemicals & Plastics
   Corp., etal., Civil Action No. 79-988), Apr. 30, 1984, Vol. I-A, 65.
 5. Faust, Charles R., Affidavit in Support of Stipulation and Judgment
   Approving Settlement Agreement (re: United States v. Hooker Chem-
   icals & Plastics Corp., et al., Civil Action No. 79-988, W.D.N. Y.,
   December20,1979), January 25, 1984, H26.
 6. Hooker  Chemicals Corporation, Occidental Petroleum Investment
   Corporation, and Occidental Petroleum Corporation, Answer to
   Amended Complaint and Affirmative Defenses, June 30,1980.
 7.  Interagency Task Force on Hazardous Wastes, Draft Report on Haz-
    ardous Waste  Disposal  in Erie and Niagara Counties, New York,
    March, 1979, pp. 111-71 -111-76.
 8. Stipulation and Judgment Approving Settlement Agreement, Adden-
    dum 1,1(B)(7)(b).
 9. Memorandum   Concerning Proposed  Stipulation   and Judgment
    Approving Settlement Agreement, 65-72.
 10. Arthur D.  Little,  Inc., S-Area Two Phase  Flow Model: Report to
    Wald, Harkrader & Ross (Washington, D.C.),  May 1983, (A.D.L.
    Ref. No. 84204-31).
11.  Stipulation   and  Judgment  Approving  Settlement  Agreement,
    Addendum V.
12.  Ibid., Addenda I-III.
13.  Ibid., Addendum IV.
 14. Ibid., Addendum I, 11(A)-(C).
15.  Ibid., Addendum I, H(A).
16.  Faust, op. cit., HI 2-24.
17.  Stipulation   and  Judgment  Approving  Settlement  Agreement,
    Addendum 1,11(B)(l)-(7).
18.  Ibid., 11(C)(l)-(2).
19.  Shifrin,  N.S.,  Affidavit in Support of  Stipulation and Judgment
    Approving Settlement Agreement, Jan. 24, 1984,1H12-14.
20.  Fogel, S., Affidavit in Support of Stipulation and Judgment Approv-
    ing Settlement Agreement, Jan. 24, 1984, Table 4.
21.  Faust, op. cit., Figure 13.
22.  Stipulation   and  Judgment  Approving  Settlement  Agreement,
    Addendum I, H1(C)(3)-(9).
23.  Ibid., 1H(C)(4)(a)-(b).
24.  Faust, op. cit., Figure 14.
25.  Stipulation   and  Judgment  Approving  Settlement  Agreement,
    Addendum I, !(D).
26.  Ibid., 11(6)-(9).
27.  Ibid., Addendum III.
28.  Ibid., Addendum II, 1(A).
29.  Ibid., Addendum 1,11(E)(l)-(23).
30.  Ibid., 1(E).
31.  Morris,  W.K., Affidavit in Support of  Stipulation and  Judgment
    Approving Settlement Agreement, Jan. 24,  1984, 525.
32.  Stipulation   and  Judgment  Approving  Settlement  Agreement,
    Addendum II, H(C)-(H).
                                                                                                       STATE PROGRAMS
                                                               531

-------
     CONSTRUCTIVE  CRITICISM ON  THE IMPLEMENTATION OF
          THE SUPERFUND  PROGRAM—A  STATE PERSPECTIVE
                                                        JIM  FRANK
                                          Hazardous Substance Control Section
                                           Division of Land  Pollution Control
                                        Illinois Environmental Protection Agency
                                                    Springfield,  Illinois
INTRODUCTION
  The purpose of this paper is to discuss selected problems states
are having with implementation of the Federal Superfund program.
The discussion will include descriptions of the problems and pro-
posed solutions from a state's perspective.  The issues to be ad-
dressed in this paper  were selected  from  three different  data
sources: a survey of all states conducted by the author and returned
in August of 1984, a survey conducted by the Association of State
and Territorial Solid Waste Management Officials between Novem-
ber, 1983  and April, 1984 and a survey conducted by the North-
east Midwest Institute  of 18 States in the Spring of 1984. These
three surveys covered a wide variety of problems. Those problems
that were prominent in all three surveys were selected for discussion.
The discussion  of the  problems and  solutions represents a syn-
thesis of the combined responses from various states.
  Before addressing specific problems, it is  appropriate to recog-
nize that this program  is relatively new compared to most federal
programs, and as such, would expect a certain number of grow-
ing pains.  Improvements have been made  in the program over
the past 14 months, and states are optimistic that the program will
continue to improve. Clearly not all of the problems are caused by
USEPA. The states can take credit for some  of the problems, such
as cumbersome procurement  procedures, a lack of  coordination
with  the Attorney General's Offices, unwillingness to commit the
necessary state resources to be a full partner in the program and  a
tendency to sit back and let the Feds run the program.
  Everyone understands that all the hazardous waste dumps were
not created  yesterday nor will they all be cleaned up tomorrow.
Hopefully, by raising these issues,  the state's administrative agen-
cies,  attorneys general, USEPA contractors, Justice Department,
the Army  Corps of Engineers and responsible parties can work to-
gether to improve and streamline the Superfund program to accom-
plish the mutual goal of cleaning  up  hazardous waste sites more
thoroughly, rapidly and cheaply than otherwise could  be accom-
plished.
  The ten issues that will be discussed in this presentation are:
•Lack of delegation of authority to regional offices
•Inability to obtain approval in a  timely manner for quality con-
 trol/quality assurance plans
•Lack of state program support
•Constantly changing guidance
•USEPA contractors and Army Corps of Engineers performance
•Contract laboratory program constipation
•Operation and maintenance costs borne solely by the states
•The unanswered question of how clean is clean enough
•Inadequacy of the hazardous ranking system
•Too much bureaucracy
                                                   DELEGATION OF AUTHORITY
                                                     The lack of delegation  of authority from headquarters to the
                                                   regions has created a giant bottleneck and  is substantially delay-
                                                   ing cleanup efforts. The following are some typical state comments
                                                   on this issue. Headquarters is trying to administer CERCLA in the
                                                   same way throughout the country. Decisions are made at too high a
                                                   level in the Agency. Anything you try that is innovative is slowed
                                                   down or blocked from approval at headquarters. If it is new, it is
                                                   bad. Headquarters staff lacks state or regional level Held exper-
                                                   ience so they play it bureaucratically safe by refusing to agree to let
                                                   states try new approaches to problem solving.
                                                     Headquarters must change its attitude and start trusting their
                                                   regions and the states more. Are so many bureaucrats needed to
                                                   review every decision? At this rate all can retire in the job 20 years
                                                   from now and never clean up a single site. Most  administrators
                                                   agree that in the formative  years of a new program a close rein must
                                                   be kept in order to ensure a desired amount of consistency. How-
                                                   ever, there comes a time to decentralize and delegate authority to
                                                   the regional offices. That time is now.
                                                     The solution to the problem is to establish a clear time table for
                                                   delegation of authority to the regions.  Areas that should be dele-
                                                   gated are: grant making authority  once a central budget has been
                                                   established; immediate and planned  removal decision-making;
                                                   selection of remedial response alternatives; decisions on cleanup
                                                   level as long as the levels are consistent with applicable guidance.
                                                   QUALITY CONTROL

                                                     The next problem is the  inability to get quality assurance/quality
                                                   control plans approved. This has been a major holdup  in pro-
                                                   ceeding with remedial investigation feasibility studies. In numerous
                                                   instances, states have had  to proceed with RI/FS without QA/QC
                                                   approvals.  In some regions of the country, there are more projects
                                                   underway  without  approved  QA/QC plans  than those with
                                                   approved plans. This is a problem irrespective of who prepares the
                                                   plan, since states  have had  as much difficulty getting plans ap-
                                                   proved as USEPA contractors and state contractors.
                                                     There are many projects on  hold throughout the country be-
                                                   cause of the delay in approving QA/QC plans. In some instances,
                                                   federal contractors have  been  told to proceed by the regional
                                                   offices without a plan approved by the regional offices. The prob-
                                                   lem has been a lack of sound specific guidance on what is expected.
                                                   Trying to get a plan approved is like a guessing game. The state
                                                   guesses what the region wants, and then the region plays Monday
                                                   morning quarterback and writes 25 page critique letters demon-
                                                   strating how much they think they know about the subject.
                                                     If this is going to be a game of nit-picking, then the States need
                                                   to know which nits the USEPA is going to pick. It is inefficient to
532
STATE PROGRAMS

-------
go through two and three rewrites of these plans, spanning from
six to 12 months, with turnaround review times taking from  one
to two months. In some cases, it appears as though the USEPA
is afraid to make a decision to approve a plan. To make matters
worse in some regions the group which approves or more appro-
priately disapproves the QA/QC plan does not work for the same
organizational unit as the Superfund staff. This sets up a situation
where the QA/QC people do not share the goals of proceeding rap-
idly with cleanups. They only worry about  100% accuracy  and
100% cost recovery.
  Everyone wants quality data which are adequately documented
for cost recovery purposes, but the tail should not wag the dog. The
solution to the problem  is to  provide  clear,  specific  guidance
followed  by workshops to disseminate  the  information to  the
states. Then the USEPA should follow the guidance in reviewing
the plans. They should place the QA/QC plan under the same ad-
ministrative head as the Superfund program in all regional offices.

STATE PROGRAM SUPPORT

  The third problem is  the lack of state  program support of  any
kind in the Superfund program. Arguments have been made  and
legal interpretations given as to why states cannot receive program
grants. Most of the arguments center on the problems of cost re-
covery or lack of authority for such grants. The states all hope the
law will be amended to specifically provide for such a grant. But if
it is not, the USEPA should rethink some of the past interpreta-
tions and policies to be more flexible on this issue.
  Following  are some problems and ridiculous situations being
created by the unwillingness to provide program support to states.
States cannot hire and  train staff until a cooperative agreement
is signed, but as soon as it is signed the state is expected to "hit the
ground running.''
  The Justice Department has asked states to forgive future over-
sight costs at  cleanup sites during consent decree negotiations be-
cause the USEPA is willing to forgive this cost and it will aid in
settlement negotiations. The USEPA then tells the state  it will cover
this cost in a continuing  cooperative agreement on the project.
Then, when the consent decree is signed by the state, the USEPA
refuses to renew the cooperative agreement for any purpose, which
leaves the state "holding the bag."
  To add insult to injury, the USEPA then uses the fund to pay its
staff and the  Army Corps of  Engineers  to do the same job at a
higher cost than it was previously paying the state. When the Army
Corps of Engineers fails to provide adequate oversight at the pro-
ject, the regional office calls the state to be updated on project
status.

Possible Solution
  The solution to this problem is for the USEPA to think about
ways  of providing sustained program support instead of finding
reasons not to. The multiside  agreement concept is a  good start,
but the prohibition on funding any state cost after a consent decree
is signed must be changed. This is a ridiculous scenario when it
plays  itself out and causes increasing overall costs to the fund. If
this latter ruling on state oversight of negotiated cleanups is  not
changed, there could be a trend for states to refuse to  sign off on
consent decrees.

CONSTANTLY CHANGING GUIDANCE
  The fourth  point concerns the problem of  constantly changing
guidance that  slows down the cleanup process by causing work to
be redone once completed. As we say about Illinois weather,  "If
you don't like the weather, wait a day; it will change." This  ob-
servation could also apply to Superfund  guidance.  The continual
shift of policies and emphasis in guidance is counterproductive; it
keeps  the entire cleanup effort shooting  at a moving target. The
solution to this problem is to not .issue  guidance until it is well
thought out. Once it is issued, make it broad enough to be flexible
so there is no need to change it for specific cases or situations.
CONTRACTOR SUPERVISION
  The fifth problem revolves around the lack of supervision of the
USEPA contractors. USEPA regional offices need to stay in closer
contact with the hired contractors including the zone contractors,
remfit contractors and the Army Corps of Engineers. Contractors
need closer supervision or regular contacts with USEPA staff.
  The current system is  set up to let contracts and not have much
more contact until the RI/FS is completed. Penalties need to be in-
cluded in contracts for late delivery of work products. The cost of
some consultants is excessive, and study costs may exceed the clean-
up  cost on some projects. The Army Corps of Engineers is not
necessarily the best group to design projects and supervise clean-
ups on all federally-led projects.
  The USEPA should have the authority to retain the Corps if it is
appropriate  and not be forced to  use them on every job.  The
Corps is not staffed in every office to handle cleanup supervision
in an on-scene coordinator role. The solution to this problem is to
provide more staff at the regional office level so that more time is
available  to  work  in  close   conjunction with the contractors
throughout the project life span. In addition, the USEPA should
not be required  to utilize the Army Corps of Engineers unless it
deems it appropriate to do so.

OVERWHELMING THE LABORATORIES
  The sixth problem is the constipation of the contract laboratory
program. This program has become bogged down and swamped
so that it is of little practical use to most states and many regions.
Long delays are encountered  in obtaining results, thus delaying
the overall rate of cleanup efforts. The solution to the problem is
to put more  resources into the program  which will cause more
organic analytical capability to become  available  in the United
States. If QA/QC plans could  be approved more rapidly for state
laboratories, more states would develop their own laboratory cap-
ability.

FUTURE O&M COSTS
  The seventh problem is that future operation and maintenance
costs are going to exceed the ability of many states to keep clean-
up systems operational and secure. There  is confusion over where
construction of the design solution stops and operation and main-
tenance begins. The USEPA  would not be favoring the lowest,
short-term cost,  "quick  fixes" which favor land disposal if they
had to operate and  maintain the constructed systems forever  like
the states will have to do.
  The USEPA would become more cognizant of the O & M costs if
required to fund them over an extended  period of time.  If the
USEPA were funding the O&M costs over the design life of the
system, more expensive short-term solutions that cost less over the
long-term would be selected.  This  change in philosophy  would
encourage more  capacity to be built into incineration, pyrolysis,
fixation and recycling systems.
  In addition,  treating groundwater would take on a more favored
status rather than placing large volumes of contaminated  soil in
vaults or providing slurry wall and clay caps for contaminated soil
and water. The solution  to this problem is for O&M costs to be
shared at the 90% federal and 10%  state  rate over the design life
of the project.

HOW CLEAN IS CLEAN?

  The eighth problem involves the uncertainty as to how clean sites
must be to be  deleted  from the national priority list. The lack of
guidance on this issue keeps the states guessing what cleanup stan-
dard to strive for. If the  USEPA has some ideas on the issue, they
should issue guidelines  on the subject. Unnecessary  delays are
caused by having headquarters second-guess the regions and the
states on cleanup levels.
  The solution is for the USEPA to publish guidelines on the most
frequently found contaminants at the NPL sites. For those com-
                                                                                               STATE PROGRAMS
                                                         533

-------
pounds where guidance is lacking, the USEPA should state the
cleanup level they prefer at  the beginning.  If they do not, they
should not require alteration of the cleanup level before the plan is
approved.

HAZARD RANKING SYSTEM
  The ranking system is the  ninth important problem. The HRS
system does not adequately consider major environmental degrada-
tion to ecologically sensitive areas, aquifers not used for water sup-
plies or threats to fish and wildlife. The HRS scoring system needs
to be more flexible to allow  major environmental degradation in
nonpopulated areas to  be dealt with by placing these sites on the
national priority  list. An alternative would be to allow each state to
place one project per year on the national priorities list even though
the project scored below the cut-off score.

BUREAUCRACY

   The last problem to  be treated is bureaucracy. The  administra-
tion of the Superfund program has become so cumbersome and so
complex that cleanup progress is being stifled. Symptoms of this
                                                       problem are slow or no decision-making, lack  of coordination
                                                       between various branches and regional offices and between reg-
                                                       ional  offices  and  headquarters,  an abundance  of constantly
                                                       changing superfluous guidance, poor coordination between the
                                                       USEPA and the Justice Department and between the USEPA and
                                                       the Army Corps of Engineers, slow turnaround times on review-
                                                       ing documents and reports  and too many people reviewing the
                                                       same reports.
                                                         The solution to this problem is to streamline processes with less
                                                       paperwork and more real work in the field. The closer to the prob-
                                                       lem the decisions are made, the faster and better the decisions will
                                                       be.
                                                       CONCLUSIONS

                                                         In  spite of all the  problems this new  and exciting Superfund
                                                       program is having, it still has the potential to be the most bene-
                                                       ficial environmental program this nation has ever seen. The author
                                                       is confident that if all try hard to communicate and cooperate, the
                                                       problems discussed in this presentation can be resolved in a timely
                                                       manner and a great program will be really Superfun(d).
534
STATE PROGRAMS

-------
       IMPLEMENTATION  OF  RCRA SECTION  3012  AT  160
     HAZARDOUS  WASTE SITES  IN WASHINGTON  STATE
                                       PATRICIA M. O'FLAHERTY
                                         RICHARD W. GREILING
                                           BARBARA J. MORSON
                                                 JRB Associates
                                             Bellevue, Washington
INTRODUCTION
  Section 3012 of RCRA provides that  all state environmental
programs develop inventories and preliminary assessments of past
or present hazardous waste storage or disposal sites. The USEPA
has designed the RCRA 3012 program so that the products of the
program (i.e., an information search  and a priority designation)
integrate with the implementation strategies under the CERCLA
program. USEPA guidance toward implementing RCRA 3012 was
published in early 1983. Concurrently, the USEPA distributed
approximately $10,000,000 to the states to fund preliminary assess-
ments at uncontrolled or potentially hazardous waste sites. Based
in part  on an inventory containing approximately 470 potential
hazardous waste sites, the Washington State Department of Ecol-
ogy (WDOE) was given USEPA grant assistance to perform RCRA
3012 preliminary assessments at no fewer than 160 sites.
  As the lead environmental protection agency in the state, the
WDOE has established and is developing a  staff expertise for the
sole purpose of implementing RCRA,  Superfund and related haz-
ardous waste programs at the state level. During the last two years,
the WDOE has surveyed hundreds of known or  suspected  waste
disposal sites, public and private generators of hazardous wastes
and those groups which collect and store, transport and treat or dis-
pose  of hazardous wastes. Through careful analysis of the  avail-
able information, the  WDOE  identified  more than 470 sites as
potentially hazardous waste sites.
  Recent legislation passed by the state of Washington provides
both  the statutory authority and an appropriation for the WDOE
to initiate site investigations and remedial responses to improper
hazardous waste management practices on those remaining Emer-
gency and Remedial Response Information System (ERRIS) or
other identified sites not presently on  the National Priorities List
(NPL) register. The WDOE is pursuing  cooperative agreements
with the USEPA to continue site investigations at a number of NPL
sites.
  In early 1984, the WDOE contracted with JRB Associates to pro-
vide technical support  in the records  search and off-site cursory
examination of 160 potential hazardous waste sites contained in
the USEPA's ERRIS files and to assist the WDOE in  the assess-
ment of these sites and determination of need for further site inves-
tigation and remediation. JRB Associates was directed to perform
RCRA 3012 Preliminary Assessments (PAs) utilizing the guidance
documents and control forms contained in USEPA Form 2070-12.
  In this paper, the authors present the  results of implementing
RCRA 3012 in the state of Washington. Key issues reviewed  in-
clude the methodologies utilized in the search for records within
multiple-tier agencies and  across a wide geographic distribution;
the difficulties encountered during the records search and PA com-
position and how they were rectified; the success employing qual-
ity control protocols to insure timely and technically consistent
site assessments; and methodologies employed in determining en-
vironmental risk for ultimate use by the WDOE and EPA in re-
ranking and updating the NPL of hazardous waste sites.

OBJECTIVES AND RATIONALE
  Under Section 3012 of RCRA, each state is required to estab-
lish a program that identifies and inventories the locations of any
site within that state that stored or disposed of hazardous wastes
(Fig. 1). Following the site's discovery process, the next step is to
complete a preliminary site assessment.  This activity involves the
collection of accessible information regarding a potential or known
hazardous waste site, using this information with a  cursory en-
vironmental characterization  to determine  the  probability  of
human health or environmental consequences  and  making an
assessment of the degree of risk and need for further actions.
  The preliminary assessment should describe information regard-
ing the site's current location and ownership; the hazardous sub-
stance's physical  state,  quantity and waste characteristics;  the
potential pollutant mobilization routes and probable targets; and
any information concerning potential or real hazardous conditions
or incidents attributable to that site. It  should also describe land
use and facility management practices  and the  potential parties
responsible for the site and its conditions.
  All of this information is presented in the Potential Hazardous
Waste Site Preliminary Assessment  form (USEPA Form 2070-12
[7-81]). Upon  review and a quality assurance control check, much
of the  data are transferred from the PA and entered  into the
USEPA's ERRIS files. The ERRIS file is the centralized data base
which tracks the history of a hazardous waste site until its final dis-
position is achieved and the site is delisted from the ERRIS register.
  The PA is primarily  intended to  be an "efficient in-office re-
view" rather than a comprehensive and complete program records
search, intensive site survey  or  field reconnaissance  effort. The
assessment confirms the site data if  they exist in the ERRIS files,
expands the existing data base with physical and cultural data and
relationships of the same with the site and, finally, provides a de-
termination if the site potentially poses a  problem. If there is a
problem, the  type and timing of followup work that should be
undertaken is delineated.
  The information gathered during this process is crucial to the ul-
timate  determination of a site's fate. The PA's conclusions will
justify  the need for  further data gathering and documentation
effort as well  as a site inspection which precedes hazard ranking
                                                                                          STATE PROGRAMS
                                                      535

-------
      CERCLA REMEDIAL
     RESPONSE PROGRAM
                           RCRA  30)2 PROGRAM
             Site
          Inspection
     National Priority List
          Nomination
                          1.   Name, Location,
                              Ownership

                          2.   Responsible Parties

                          3.   Brief Site Description

                          1.   Ha/ardous Substance
                              Information

                          5,   Environmental
                              Descriptions

                          6.   Demographics

                          7.   Land INr Information

                          8.   Pollution Mobilisation
                              Routes and Targets

                          9.   Priority Designation

                          10.   Conclusions and
                              Recommendations


                          EPA  Form 2070- 12
        Remedial Plans
         and Actions
                          Figure 1
      Integration of RCRA §3012 Program with Implementation
               Strategies of the CERCLA Program

and NPL nomination. In some cases, the PA may go beyond an
initial screening and actually provide enough documented informa-
tion to permit a Hazard Ranking System (HRS) score which further
evaluates the magnitude or severity of that site's hazardous wastes
problem.  Based  on the assigned value,  a  site  may  then be
nominated for NPL inclusion.
  In consideration of the above, it is clearly evident  that despite
the cursory nature of the PA's data gathering  efforts, this phase
within the CERCLA remedial response program represents a decis-
ive and significant step in the recognition, evaluation and ultimate
mitigation or correction of hazardous waste disposal problems.

METHODOLOGIES

  JRB Associates approached the performance of all project tasks
within an organized and sequential chain of activities (Fig. 2). The
RCRA 3012 preliminary assessments  were performed in  four
phases:
•Phase 1-Project definition
•Phase 2-Records and information search
•Phase 3-Correction of data deficiencies and assemblage of the pre-
 liminary assessment
•Phase 4-Review and PA finalization
  Each of these phases provided for  a  timely flow and thorough
performance of the overall assignment.

Project Definition

  In Phase 1, Project Definition (Fig. 2a), coordination with the
WDOE headquarters and its four regional offices  was initiated  in
order to establish proper protocols and make key  staff contacts.
This action also familiarized JRB staff  with state files, programs,
records, indexes and resource locations. During this  initial coor-
dination, JRB was able to determine the format and material types
necessary for data base deliverables that  would fulfill the needs and
requirements for  WDOE  computer/word  processing  systems.
Finally, contacts were arranged with appropriate federal and local
agencies such as USEPA-Region X and county health departments
to establish similar coordination and records access.
  Upon completion of this groundwork,  JRB and WDOE staff
finalized a performance schedule that  would satisfy the state's re-
quirements. Flexibility was maintained within this schedule by
batching PAs into groups of approximately 30 to 35 sites with in-
terim deadlines in order to achieve concurrent draft PA review and
quality  control with the state  personnel. This also allowed the
WDOE to make any site substitutions  on the latter batches should
pertinent information surface regarding any new site(s) that would
require immediate attention. Finally, the selection of JRB project
team leaders and  specific site assignments set the stage for site in-
formation collection.

Records and Information Search
  During Phase 2 (Fig. 2b), the Records and Information Search,
all levels of government (federal, state and local) were targeted for
each site's  investigation.  Private agencies  contacted were usually
limited to research or consulting firms  that had been contracted by
regulatory or public agencies to perform pertinent hazardous waste
investigations  and to private haulers of  solid or  liquid wastes
known to  or suspected of hauling wastes from or to a particular
site. The information searches undertaken included reviewing exist-
ing agency files; communicating  with  knowledgeable staff; re-
searching ongoing studies; reviewing archives or historical records;
collecting technical references; visiting local and regional libraries;
and compiling  media releases.  Specific information sources for
completing the 2070-12 form included:
Waste Types and Quantities—Because of the specificity of the in-
formation, agency  records,  ongoing studies  or familiar  staff
appeared to be the best sources. Examples of technical references
used included:
•NFPA's Fire Protection Guide
•Sax's Dangerous Properties of Industrial Wastes
•NTIS'  Water Related Environmental Fate of 129 Priority Pollu-
 tants
•Brown's Ecology of Pesticides
•NOAA's Puget Sound Marine Ecosystem Analyses
Geology and Hydrology—Sources included:
•U.S. Geological Survey
•Geohydrologic investigations and reports
•Well logs
•Water supply bulletins and monographs
•Washington stream catalogs
•USGS stream flow data
•Physiographic and topographic maps
Natural Resources—Sources included:
•U.S. Fish and Wildlife Service
•Washington Department of Natural Resources
•Washington Department of Game
•Washington Department of Fisheries
•Washington Natural Heritage Program
•National Wetlands Inventory
•National Weather Service
•Regional air pollution control authorities
•Local regulatory and conservation organizations

Social and Human Resources—Sources included:
•Washington Department of Social and Health Services
•Local county health departments
•Local water and sewer districts
•Public utilities and  irrigation districts
•Federal, state and local census data
•Puget Sound Council of Governments
•Business license maps and records
•Flood Insurance Rate Maps (FIRM)
•Tax Assessments
•State archives
536
STATE PROGRAMS

-------
                                Phase I - PROJECT DEFINITION
    Obtain WDOE Staff Contact
 Register, $3012 site File Index,
  Records Location! and Access
          Procedures
                                          Initiate Coordination with
                                       WDOE HQ and Regional Offices
                                      Identify Needs and Data  I/O
                                        Requirements for WDOE
                                      Computer/Word Processing
                                              Sy items
                                       Initiate Coordination with EPA-X,
                                       County Health Departments, and
                                          Other Appropriate Public
                                         Agencies for Contacts and
                                              Records Access
                                       • Finalize Performance Schedule
                                        and Site Batch Organization
                                        with WDOE.
                                       • Selection of JRB PA Team
                                        Leaders; Complete Specific
                                        Site Assignments.
                                           (go to  Phase II)
                                              Figure 2a
           Performance of RCRA Section 3012 Preliminary Assessments for WDOE
                          Phase  11    RECORDS INFORMATION SEARCH
 Coordinate with Local Land Use,
    Air, and Water Resource
      Management Agencies
 Review State and Local Archives
 for Historical Ownership Records
 and General  Descriptions of the
   Site's Neighborhood During
       Period of Operation
 Identify Elements of the Physical
 and Human Environment Poten-
   tially Affected by the Site:

e Air, Soli and Water
• Onslte Worker, General Public
• Flora, Fauna,  Food Chain
• Offslte Migration and Damage
                                         On a Batch Basis,  Perform
                                        Records Search at WDOE and
                                           EPA Offices.  Verify
                                          Completeness of Agency
                                          Records and 2070 Part I
                                           Coordinate with Local
                                        Municipal Agencies, Utilities,
                                           and Legal Authorities
    Identify Feedstocks and
 Quantities, and Waste Type*
  and Quantities Reportedly
 Stored, Treated or Disposed
 on Site.  Characterize Waste
 Types, Physical Properties,
       •nd TSD Status
 Review Technical Information
Available In the Literature such
     as SAX, NFPA, etc.
                                        Establish Flies for Each Site
                                        Replete with Photocopies of
                                     Relevant Records and Documenta-
                                         tion of AM In-person and
                                            Telephone Contacts
                                      Perform Records Search of
                                      WDOE and Local Well Logs
                                      Within Three-Wile Radius
                                            of §3012 Site
                                    Review USGS, WDOE, W5DOT
                                      and SCS Soil Reports and
                                      Hydrogeologlc Summaries
  Secure/Review NWS and Local
  Meteorological Data, Plus HUD
Flood  Insurance Site Designations
Identify Probable Regional E Site
Ceohydrology; Surface & Ground-
  water Hydrology; Vectors for
 Hazard Migration; Contamination
     of Air, Water and Soil
                                         Review Files for Degree of
                                              Completeness
                                          (go  to  Phase III)


                                              Figure 2b
           Performance of RCRA Section 3012 Preliminary Assessments for WDOE
                                                                                                 STATE PROGRAMS
                                                                                                                                      537

-------
       Phase 111  -  DATA DEFICIENCY CORRECTION

          AND PREPARATION OK FORM 2070-12
                                YES
                                               NO
                                     Identify Dm Caps ind
                                     Information Needs ind
                                     Poiilblc Saurcei of Dels
        Mar* Rigorous
        Records Search
                                               NO
                                    Perform Off-site Windshield
                                    Survey for Data Collection/
                                   Con firm* I ton end Prepare Site
                                  Survey Memo with Location Map
                                  • Assemble All Data on EPA
                                    Form 2070-12.
                                  • Assign Priority Assessment
                                    Ranking.
                                  • Enter on JRB Computer!led
                                    Data Base Management System.
                                    In-House QA/OC Review of
                                    Form 2070-12 and Data Base
                                   Output by Team Leaders and
                                        Project Manager
                                      (go to Phase IV)
                           Figure 2c
 Performance of RCRA Section 3012 Preliminary Assessments for WDOE


  Files were established for each site as information was generated,
and photocopies of records and documentation of all telephone
and in-person contacts were maintained to insure record verifica-
tion and to facilitate the inclusion of these data on the PA. A com-
pleted bibliography and contact summaries with all appropriate in-
formation sources were maintained and updated for use as a  final
deliverable.

Correction of Data Deficiencies

  Any data  gaps or  deficiencies were rectified during Phase  3
(Fig. 2c) if that information was reasonably accessible or if the
team leader determined that the budget and time constraints  were
not strained by a more rigorous investigation (20 to 40 hr has  been
estimated as the appropriate range of time required to fully com-
plete a PA). Off-site or "windshield surveys" were also imple-
mented during this phase. They were extremely useful for deter-
mining the current status of a site as well as collecting additional in-
formation regarding its environment and  its potential risks. A
site survey  memorandum, including travel directions, location
                                                        maps and site sketch, was added to the main site file record follow-
                                                        ing all "windshield surveys,"
                                                          At the conclusion of the records search and site survey, the 2070-
                                                        12 form was prepared and began a rigorous in-house quality con-
                                                        trol review. To ensure compatibility and  accuracy among the site
                                                        assessments,  the  task of reviewing  and adjusting the  PA  was
                                                        assigned to the team leaders and project  manager, who evaluated
                                                        each assessment for overall quality and completeness.  Guidelines
                                                        on PA priority rankings were provided by the WDOE  and are as
                                                        follows:
                                                        •High—Imminent health or major environmental threat  highly sus-
                                                         pected
                                                        •Medium—A. site is highly suspected to present a potential prob-
                                                         lem; evidence from  sampling, direct observation by  regulatory
                                                         agency or history of problems at site
                                                        •Low—Unresolved  question  but not  highly  suspected;  alleged
                                                         problem (tip from employee or member of the public) or unknown
                                                         how facility disposed of suspected hazardous waste
                                                        •None—No evidence to suspect a problem
                                                          During this time, the form was also entered onto a data base
                                                        management  system, utilizing JRB's inhouse microcomputer, an
                                                        IBM PC. This data base management system,  recording all pre-
                                                        liminary assessments, is compatible with WDOE's data  base man-
                                                        agement system. This computerized information management sys-
                                                        tem was prepared specifically to accept the RCRA 3012 data base.
                                                        Review and PA Finalizalion
                                                          Upon a final satisfactory review by the team leaders and a de-
                                                        termination by the JRB project manager  that the PA satisfied in-
                                                        ternal quality control standards, the PA was ready for submission
                                                        to the WDOE.
                                                          When the batch  of site  assessments had been compiled, they
                                                        were organized and  identified by priority ranking and level of risk.
                                                        At this point, they were transmitted to the WDOE for  Phase 4
                                                        (Fig. 2d). A  report summarizing all  of  the assessments accom-
                                                        panied this deliverable which briefly highlighted the categories of
                                                                         Phase /V   REVIEW AND PA FINALIZATION
Assemble ell Site Assessments
and Identify by Priority Ranking
the Level of Risk and WDOE
Action ; Assign Levels of
Action from Emergency
Response to Deflating
'

Prepare e Draft Summary Report
of All 13012 PA Activities
'

Submit to WDOE for Review
and Comment
<

Meet with WDOE end Respond to
all WDOE and Inter-agency
Comments; Prepare and Produce
Final Form 2070-12 (PA)
<

Project Completion; Submission
of 160 Preliminary Assessments.
Final Summary Report,
Data Base System and Compiler,
Master Site Files and Index, and
Completed Bibliography end
Reference Index
                                                                                  Figure 2d
                                                         Performance of RCRA Section 3012 Preliminary Assessments for WDOE
538
STATE PROGRAMS

-------
waste sites,  clarified the priority ranking and identified problem
areas or persistent  data  gaps. This report also summarized JRB
recommendations to remedy informational needs and presented
technical guidance for possible remedial or mitigative actions. The
WDOE, therefore, had sufficient information to pursue its own re-
view and edit  each potential hazardous waste site.  Concurrent
with the state's review, JRB commenced work on the next batch of
site records search and assessments.
  A meeting with WDOE's project manager and staff to review
all WDOE edits and questions was the final step in Phase 4. The
PA then underwent a final revision and was resubmitted to WDOE
in a final form with backup copies. At the conclusion of all 160
preliminary assessments, a final summary report,  the  data base
system with a compiler, the master site files and index and a com-
pleted contract summary and bibliography were also submitted.

RESULTS AND DISCUSSION

  After completing 160  preliminary assessments  for the WDOE,
the staff at both JRB and WDOE determined that the amount of
time predicted  to complete average site  assessments was  approx-
imately 24 working hours, well within the original time estimate.
JRB also believed it achieved great success in employing  quality
control protocols that resulted in timely deliverables and  quality
products. Diligence in maintaining strict documentation protocol
was well rewarded during state and inter-agency reviews. JRB can-
not overemphasize  the importance of maintaining precise records
of information sources.
   Some aspects of the PA process were fraught  with difficulties,
and the following discussion elaborates the source of problems as
well as the methods JRB utilized to overcome them.
   The culmination of the records search, file documentation, site
surveys and the interpretation of all the above optimally leads to  a
completed PA. The completed Form  2070-12 should  contain
enough information pertaining to a potential site's risks as well as
its environment to permit the reviewer to assign a rank of high to
none indicating the degree of need for further site investigation
and monitoring activities.  Based on this ranking, the concerned
agency can then begin  remedial action. The first difficulty  en-
countered in this process was interpreting each ranking designa-
tion. Fortunately for the JRB  staff, WDOE provided guidelines
for designating a ranking.  However, it would also have been use-
ful if specific regional examples had been provided by the USEPA
that could have facilitated the initial scoring process. The necessity
to produce a standard for assessing potential hazardous waste sites
cannot be disputed, but the variability within a region and the de-
gree of concern by local regulatory agencies can  often complicate
the overall objective.
   Based  on the  completion of  over  160 PAs for the WDOE,
approximately  30% of the sites had sufficient information available
in various existing files or records systems to  assemble a com-
pleted Form 2070-12. An additional 20% required information to
complete assessments;  the information was  easily obtained by
verbally contacting WDOE inspectors and local health or public
works  officials. Cooperation from these various  agencies was not
only gratifying, but also very illuminating because local author-
ities often had ample familiarity with a potential site due to their
proximity to it. For the  remaining 50%  of the sites initially nom-
inated to receive the assessment, either very limited  information
was available regarding the final fate of hazardous wastes or the
sites were often no longer in operation with few or no records re-
garding the types  of wastes  generated or handled. For example,
small metal plating and fabricating industries were sites for which
there was often little file information regarding the final destina-
tion of their dragout sludges. In all of these cases, a 2070-12 form
was completed for each site with JRB's recommendations for con-
tacting the owner or operator and/or a detailed site investigation.
  A common  difficulty encountered related to sites  that were
selected for review  based only on rumors or complaints and with-
out any additional background information that would allow  a
complete assessment. In these cases, the records search  revealed
nothing more than an  initial complaint  record  with  perhaps a
subsequent site inspection that could not resolve or identify a spe-
cific problem. This  situation, of sourse, is not unusual and it
proved to be challenging to  JRB staff. Essentially, the  reviewer
sharpens his or her investigative skills to provide, at a minimum,
a description of the site's environment in order  to evaluate tar-
gets and pathways for potentially hazardous substances.
  A recurrent dilemma was assessing the potential consequences of
abandoned landfills. In most instances, these sites were public land-
fills that were not regulated for the types of wastes received. Others
were industrial or otherwise private on-site  dumping  areas that
again may have been inactive and even covered  over  but would
have logically received process waste materials that may or may not
have had hazardous  components. Oftentimes records were re-
viewed which documented that landfill leachates were sampled for
conventional pollutants  (e.g., TOC, pH,  COD, BOD, nitrogen,
chlorides and specific conductance) and perhaps some of the heavy
metals, but not other priority pollutants or hazardous wastes that
conceivably could be present.  In these cases, it became impor-
tant to emphasize the hydrological and geological  characteristics
in order to evaluate risk potential. Where no records existed con-
cerning the types of wastes present, JRB searched local historical
resources as well as the state archives in order to provide a descrip-
tion of that site's environment wherein potential waste sources may
be located. Old tax assessments and property locator  maps  were the
best sources of this information.
  Some sites did not have hazardous waste problems and thus were
inappropriate for RCRA or  CERCLA response. These frequently
were the more classical water quality problems  from biological
wastes and included farming and rendering activities. While  these
problems  are to be excluded from  further RCRA action, it does
not necessarily mean that there are no human health or  environ-
mental risks. Based  on this  analysis, JRB  completed the Form
2070-12 with as much information as could be derived  and docu-
mented the presence (or absence) of a problem and the degree of
environmental or  population risks. This  step provided a useful
service to the contracting agency that  is responsible for  a wider
range of environmental problems than only hazardous materials or
wastes.
  JRB encountered several areas of difficulty when completing the
USEPA Form 2070-12 that required interaction with the WDOE
and USEPA-Region  X  in order  to  overcome. A summary of the
major difficulties encountered are presented below, and a copy of
USEPA Form 2070-12 is presented in Figure 3 with  those sections
highlighted.
Responsibilities (Part 1,  Part III)
  The 20-70-12  form does not  provide adequate space  for dis-
tinguishing between past and present owners and operators. Many
sites have changed ownership and land usage several times. A pro-
vision within the form to enumerate past responsible parties would
not only be useful in understanding the hazardous waste site's his-
tory, but it could also reduce  the amount of future time spent con-
ducting responsible party searches.

Waste States, Quantities and Characteristics (Page 2, Part II)
  Quantities of wastes  are  designated  in three categories:  tons,
cubic yards or number of drums. If the site involves  a pool of con-
taminated materials, it would appear that  a liquid measurement
such as gallons or liters would provide a clearer picture of that site.
This would provide a future site investigation team with a better
idea of what to expect before arriving at a facility.

Hazardous Substances (Page 2, Part IV)

  The  chemical concentrations  of hazardous wastes can be re-
ported in  several ways.  The form should  distinguish between the
results of environmental release analyses from the analyses of the
waste materials exclusively. JRB assigned sub-headings within this
section to  allow for these differences.  Chemical Abstracts  Serv-
ices (CAS) numbers should not be included if the possibility exists
                                                                                                 STATE PROGRAMS
                                                           539

-------
                          xvEPA
                                        POTENTIAL HAZARDOUS WASTE SITE
                                            PRELIMINARY ASSESSMENT
                                       PART 1 • SITE INFORMATION AND ASSESSMENT
                                                                                       I IOCNTVICATIOM
01 3!»Tf|0}*lf ft
                         II- SITE NAME AND LOCATION
                                                                OJ*'*l|i HOu'l •*! 0*
                              IIIS L>,|,UW
                                                                *I«>II|M»CBU
                         [V CHARACTERIZATION OF POTENTIAL HAZARD
                           „ YES  DATE
                           '.: NO
                                        : A epfc      a EPA CONTHACTO«    ,  c v*ri
                                        ; t LOCit HE*ifH OFFICIAL  1 F OTMtB 	
                                                                                     - 0 OT>«»» CON T
                                                  CONTRACTOR NAMElSI	
                            » ACHVE  ~l B INACTTVt  C C
                          PRIOKITT ASSESSMENT
                         I. INf ORMATION AVAILABLE FROM
                                                             Figure 3
                                                     EPA Form 2070-12. Page I
for chemical  reactions among  hazardous  materials or  between
wastes and natural elements. The assessor may not know what the
altered chemical states are  without  precise analytical data. Con-
firmed environmental contamination at some sites was difficult to
establish from background levels if the contaminants were similar
to elements found naturally. An example of this situation was land-
fill leachate with low levels of heavy metals.  Inadequate data on
these background levels preclude adequate assessment of risk  and
judgment that contaminants  were above the natural concentra-
tions.
                                                         Hazardous Conditions and Incidents (Page 3, Part II)

                                                           Site specific geologic and hydrologic data are an important fac-
                                                         tor  in determining risk assessments. Geology in  the mountainous
                                                         regions of Washington is quite diverse. Attempting to define char-
                                                         acteristics of the aquifer of  concern and its appropriate strati-
                                                         graphic sequence may not always be precise, particularly in remote
                                                         or unpopulated regions if there are no well logs or specific geo-
                                                         logical information available. Furthermore, defining the aquifer of
                                                         concern  in  northwestern Washington  in areas not serviced  by
                                                         municipal supplies may not be a simple task. Water bearing zones
 540
STATE PROGRAMS

-------
_ _._- POTENTIAL HAZARDOUS WASTE SITE
C>FPA PRELIMINARY ASSESSMENT
^^fc-l ** PART 2 -WASTE INFORMATION
(.IDENTIFICATION
l"oi STATE

02 SITE NUMBER

II. WASTE STATES, QUANTITIES, AND CHARACTERISTICS
01 PHYSICAL STATES ,c*w» -m« «*>v « WMT1 OUAlfflTT AT *RI.
A SOLID E SLURRY "***MW»«**«#
C SLUDGE
0 OTHER
Q GAS f ,
-~..,., : MWau^ ,.-rnil 	 ',
A TOXC E SOLUBLE
B CORROSIVE F INFECTIOUS
C RADIOACTIVE 0 FLAMMABLE
D PERS5TENT H WRITABLE
HOHIY VOLATILE
EXPLOSIVE
REACT VE
INCOMPATIBLE
,4 NOT APPLICABLE
III. WASTE TYPE
CATEGORY
SLU
OLW
SOL
PSD
OCC
IOC
ACD
BAS
MES
SUBSTANCE NAME
SLUDGE
OILY WASTE
SOLVENTS
PESTICIDES
OTHER ORGANIC CHEMICALS
INORGANIC CHEMICALS
ACIDS
BASES
HEAVY METALS
01 GROSS AMOUNT









02UN1TOFMEASUH









O3 COMMENTS











IV HAifAftOOm"WMTftHCP>[iiiiaMiiiJtif*-M^n'i4ij-i4'-iMf nri tut ' " ' •
oiCATOQW
















*, - «suwwce«A«
















cacAflMUM«*





^










WHTWM^MMTOUIMITMOO
















WCONUNTKAT1GN
















^N^JrfiSiSw
















V. FEEDSTOCKS ,s« -«*•». MCM **»•».
CATEGORY
FDS
FDS
FOS
FDS
VI. SOURCES OF
01 FEEDSTOCK NAME




02 CAS NUMBER




CATEGORY
FDS
FDS
FDS
FOS
01 FEEDSTOCK NAME




02 CAS NUMBER




NFORMATION tCJ. w«*e r,f..«K« *g .«."*. .mwtwnu .«KW. I

                      EPA FOflM 3070 12 (7 81J
                                                          Figure 3
                                                   EPA Form 2070-12, Page 2
are numerous and distributed throughout the stratigraphic column.
Depending on personal preference and expense, private wells can,
and do, tap a  multiple of water-bearing strata. Compensation for
this problem primarily depended upon the technical knowledge of
JRB staff geologists as well as local water purveyors.
  Determination of the population potentially affected for poten-
tial surface, groundwater, fire or air contamination is based on the
HRS guidelines of a one  through four  mile radius.  Using these
specified distances required apportioning census tracts or blocks to
conform to that area. This apportionment may not always conform
to true population densities and clusters.
Continuation of Hazardous Conditions and Incidents (Page 4)

  USEPA guidelines indicate that contamination of food chain
applies to human consumption of food/forage crops that may be
grown on contaminated grounds or are irrigated by contaminated
surface and/or groundwater. This section has also been interpreted
to mean environmental and biological food chain contamination, a
much more  complex issue. Predicting or determining biological
food  chain risks  would require a consideration of trophic rela-
tionships as well as an understanding of the physiology and com-
pensatory characteristics of many organisms. Thus, it may be wiser
to avoid speculation of biological  food chain contamination on a
                                                                                               STATE PROGRAMS
                                                          541

-------
vvEPA
^^^' ** PABT3-DESC
POTENTIAL HAZARDOUS WASTE SITE '
PRELIMINARY ASSESSMENT "'

•XNWCATIOtt


&HAZAMXMMCONMTWNIANOMCmNTI . . * . •',
01 A QROUNOWATERCONTAMINATON

oi e SURFACE wAUHCONTAMiNAtiON

01 C CONTAMnATON Of AM
03 POPULATION PQTlNtiAu > ArffCttD
01 0 f«t EKPiOWVt CONDITIONS
03 POPUIATON POTENTIALLY AfrtCTtO
01 « 0«ŁCT CONTACT
03 POP\JLATON POTENTIALLY ATFECTEO
01 f CONTAMNATONOF SOU
•JU
01 G O**WKiNGWAlŁflCONtAM*,»HON
03 POPUIATOM POTENTIAL » AfMC'CD
Oi .. h wOMKEt EHPOSUME vuufo
OJ WOBHERSPOTtNTUU.* AfFECTtD 	
01 i_ 1 POPULATION ElPOSURC »*JUHV
03 POPULATION POTENTIALLY AFFECTED

01 OSSIRvtOiOAIt PO'fNTXAL ALUOfO
04 NAfWAfrvE Of SC^rf* "ON
0? OttSt ffvt 0 ' OA It > <*OT( NTiAi AixEQCO
0< MARMAtlVf OtSCWO*
07  «T(
04 NAMBAtM 0*SC(W>f«:>,

OJ OHMWVtO D*'( « *OT*N1Ui AUJOCO
U* NAflMATlVf M*C«WTl>»
02 OKtERVEOiOAri PC'I
02 C O«S*«V€OiOAH 	 , I^J»|

07 OftSCftVED'DAiC < *OT(
02; o*s««veo'OATc ._, 	 	 	 s »o't'
O2 OBSCRvCDiDATf . POTT*




«n*i AU.EOCD

Tvk. ..' ALLCOCO
                                                           Figure 3
                                                   EPA Form 2070-12, Page 3
site-specific basis and report the potential for food chain transfer
and known bioconcentration factors of relevant contaminants that
are found in the literature.
  Contamination of sewers presented another problem to the JRB
staff because of the large numbers of districts in some Washington
counties. It was often a tedious task locating potentially affected
sewer systems and determining the potential for migration risks.

CONCLUSIONS

  Based  on the completion of  160  preliminary  assessments in
Washington State, JRB Associates can make the  following con-
clusions:
•The PA process, which  encompasses the records search to the
 final assemblage of the 2070-12 form, can be accomplished in an
 average of 24 hr.
                                                       •Documentation of all information sources is crucial to supporting
                                                        the final PA ranking and supporting any future actions.
                                                       •The easiest data to find in most cases included the description
                                                        of demography and environmental characteristics such as hydrol-
                                                        ogy,  geology, floral and faunal characteristics. Environmental
                                                        characteristics regarding sites located in remote areas, however,
                                                        were not always site specific or as complete, and in these cases
                                                        relevant technical judgments were called upon to provide ade-
                                                        quate descriptions.
                                                       •The most  difficult information to find was the hazardous waste
                                                        site data such as  the types and quantities of materials disposed
                                                        and the physical characteristics of the site. This was difficult be-
                                                        cause, in many cases, these sites had ceased operation long ago or
                                                        had altered their  practices and,  consequently,  no records were
542
STATE PROGRAMS

-------
                                                    POTENTIAL HAZARDOUS WASTE SITE
                                                       PRELIMINARY ASSESSMENT
                                          PART3-DESCRIPTION OF HAZARDOUS CONDITIONS AND INCIDENTS
                                                                                            I. IDENTIFICATION
                             01 STATtloj Sn* NUMBER
                           01 n J DAMAGE TO FLORA
                           04 NARRATIVE DESCRIPTION
                                                               02 I , OBSERVED (DATE	I   G POTENTIAL    L] ALLEGED
                           01 ^ K DAMAGE TO FAUNA
                           04 NARRATIVE DESCRIPTION ,~cw.,,.-..,
                                                               02 :. OBSERVED IDATE
                                                                                   	I   O POTENTIAL    U ALLEGED
                                  TAMINATION OF FOOD CHAIN
                                           ^^^
                                                               02 ' : OBSERVED IDATE	
                                                                                         ~ POTENTIAL    (" ALLEGED
                           01 d M UNSTABLE CONTAINMENT OF WASTES

                           03 POPULATION POTENTIALLY AFFECTED	
02 : OBSERVED IDATE 	

04 NARRATIVE DESCRIPTION
                                                                                         ~, POTENTIAL    H ALLEGED
                           01 G N DAMAGE TO OFFSITE PROPERTY
                           04 NARRATIVE DESCRIPTION
                                                               02 '~ OBSERVED (DATE
                                                                                     _ I   C POTENTIAL     ;: ALLEGED
                           Ol^^OCONTAMINATIONOF^^WFRS^^TI^RMORAI
                                                       INS VWVTP? 02 : OBSERVED (DATE
                                                                                     _ )   -  POTENTIAL
                           01 I P ILLEGAL-UNAUTHORIZED DUMPING
                           04 NARRATIVE DESCRIPTION
                                                               02 ': OBSERVED IDATE
                                                                                   	I   ^ POTENTIAL
                           05 DESCRIPTION OF ANY OTHER KNOWN. POTENTIAL OR ALLEGED HAZARDS
                          III. TOTAL POPULATION POTENTIALLY AFFECTED: .
                          IV. COMMENTS
                          V. SOURCES OF INFORMATION
                         EPA FORM 2070 12(7 811
                                                                 Figure 3
                                                        EPA Form 2070-12, Page 4
 available to reflect these past practices, number of workers or in-
 formation  regarding access to the site.  Whenever this situation
 occurred, it was imperative to provide, at a minimum a descrip-
 tion of the site's environment and, particularly, a historical over-
 view of the types  of adjacent land use that was prevalent at the
 time of the alleged or known hazardous wastes handling. A search
 through archival records of tax assessments and county property
 maps provided the most valuable source of this information.
  The performance of  the  preliminary assessments provided the
WDOE with a completed Form 2070-12 and associated files for 160
potentially hazardous waste sites, and,  more importantly, it pro-
vided the state with an awareness of these sites'  relative degree of
hazard to populations at risk and to the environment. Overall, this
        awareness is the most beneficial  result of  preparing preliminary
        assessments because the process is designed to screen and indicate
        potential problem sites and can assist the state's determination of
        the next appropriate action including site investigation, MRS scor-
        ing, nomination for NPL or even delisting.


        ACKNOWLEDGEMENTS

           This project was funded by the Washington State Department
        of Ecology under Contract 0C-84094. All WDOE regional offices
        assisted with the records gathering effort and guidance was pro-
        vided by Ned Therien, WDOE RCRA Section 3012 Project Man-
        ager.
                                                                                                         STATE PROGRAMS       543

-------
              RCRA 3012 AND SUPERFUND  ENFORCEMENT
                                     AT THE  STATE LEVEL

                                             CHARLES R. FAULDS
                                            DANIEL L. SCHEPPERS
                                     Texas Department of Water Resources
                                                    Austin, Texas
                                                DAVID  JOHNSON
                                                Engineering-Science
                                                    Austin, Texas
THE TEXAS PROGRAM

  In an effort to define the number of sites where hazardous sub-
stances may  have been handled, spilled or dumped, the USEPA
set up the Emergency and Remedial Response Information System
(ERRIS). At  present, over 16,000 potential hazardous sites are rep-
resented on  the national ERRIS list, and the USEPA estimates
that perhaps 22,000 sites ultimately will be identified. The Agen-
cy's fiscal  1983 appropriations bill provided $10 million from the
Superfund to perform activities authorized under Section 3012 of
RCRA. The  bill provided for one-time grants to  assist states in
completing the identification and evaluation of potential hazardous
waste sites. The grant allocation for Texas was $677,000 based
upon 1,109 sites.
  The  evaluation process  for sites  on the ERRIS list involves a
sequence of  investigations. The  first step is a preliminary assess-
ment which  is  the collection  and  review of readily  available in-
formation. Second, if the preliminary assessment determines that a
potential hazard exists, a Site Inspection is scheduled. The Site In-
spection can lead to additional investigation, emergency actions,
enforcement  action or ranking of the site for inclusion on the Na-
tional Priorities List (NPL). In Texas, because of the large number
of sites, the state and  the USEPA  have divided the task of inves-
tigating ERRIS sites.
  The USEPA's Field Investigation Team (FIT) has evaluated ap-
proximately 580 sites on the Texas  ERRIS list. The State of Texas
is responsible for the  remaining 527 sites. Prior to this coopera-
tive agreement, the state had completed investigations on 264 sites;
under the RCRA 3012 Program, the State will complete inspections
of the remaining 265 sites in 1984.

Site Discovery and Preliminary Assessments

  The sites presently listed in ERRIS have been identified through
a discovery  process. They may have been identified under the
RCRA 3012  program or by other  means including  citizen com-
plaints, existing site notification programs such  as RCRA and
CERCLA, land use records, aerial photo-imagery or federal, state
or local governmental  records. The  ERRIS list includes only those
sites which have been reported to the USEPA and is not a compre-
hensive listing of all potential hazardous waste sites. Sites will con-
tinue to be added to the ERRIS list as they are discovered.
  Following  site discovery, a preliminary assessment  is performed
to determine if further action at the site is required. In many cases,
the preliminary assessment and site identification activities are
combined. The preliminary assessment is a quick, low-cost char-
acterization of the site,  including  determination of the potential
presence of hazardous materials, past and present facility waste
management practices,  pathways  allowing off-site  transport  of
waste materials, potentially impacted populations and responsible
parties.
  This preliminary evaluation involves review of readily accessible
data on the site—usually from federal, state or local governmental
agencies. Additional information on land use, flood hazard poten-
tial, groundwater resources and meteorological characteristics is
typically compiled and reviewed for the preliminary assessment. In
                                                     some cases, off-site reconnaissance is necessary to provide addi-
                                                     tional information.
                                                       The completed preliminary assessment can result in one of three
                                                     actions. In certain cases, where an imminent threat to the environ-
                                                     ment is  identified,  emergency response  may be warranted,   in
                                                     other cases, the preliminary assessment may result in a recommen-
                                                     dation of no further action. However, in most cases, the site will
                                                     be scheduled for a site inspection to more completely identify po-
                                                     tential hazards and to obtain additional information.
                                                       A total of 177 preliminary assessments were performed under the
                                                     Texas RCRA 3012 program. Of these, none were identified as re-
                                                     quiring immediate emergency response.
                                                       A total of 82 of these sites were found to have no on-site haz-
                                                     ardous waste handling or disposal or were determined to pose no
                                                     hazards to the environment. Many of these sites were municipal
                                                     landfills used for domestic waste disposal only or sites which  had
                                                     already been closed under  TDWR  supervision. Trash and waste
                                                     haulers with no on-site waste handling, who disposed of wastes in
                                                     landfills already listed  in ERRIS, were also identified in this group.
                                                     Off-site reconnaissance was conducted at a large majority of these
                                                     sites and was invaluable in identifying sites where no further ac-
                                                     tion was necessary. The remaining 95 sites were recommended for
                                                     on-site inspections to further delineate potential hazards.

                                                     Site Inspections
                                                       The purpose of a site inspection is to characterize potential haz-
                                                     ardous waste sites by providing a data base sufficient to screen out
                                                     sites which will not be a problem, provide additional information
                                                     for state enforcement activities or complete the hazard ranking sys-
                                                     tem (HRS) on sites that do pose a problem.
                                                       Objectives of a site inspection include the confirmation of pre-
                                                     liminary assessment data, development of data unavailable during
                                                     the preliminary assessment and an update on site conditions if there
                                                     are indications that undocumented changes may have occurred.
                                                       It is important to provide analytical evidence that a legitimate
                                                     hazardous waste control problem exists. The site inspection accom-
                                                     plishes this through sampling both on-site  and off-site to obtain
                                                     evidence  that hazardous  materials are  present on-site and to
                                                     demonstrate the migration of such materials off-site.  These studies
                                                     are limited efforts and are not intended to take the place of a com-
                                                     prehensive field investigation.
                                                       The site inspection  is not an intensive, complete environmental
                                                     assessment. Rather, it is a quickly implemented investigative effort
                                                     limited to gathering data. Site inspection activities are restricted to:
                                                     field  measurement of ambient conditions,  the documentation of
                                                     observations regarding hazard conditions at the site and at sample
                                                     collection locations and the collection of grab  samples including
                                                     samples  from on-site  soils,  waste spillage, open waste containers,
                                                     waste pits and lagoons, off-site soils, surface waters and ground-
                                                     waters.
                                                       Specifically excluded from site inspection activities are geophysi-
                                                     cal testing, groundwater monitoring well installations and all other
                                                     activities that require  detailed prestudy or specialized techniques.
                                                     Finally, studies aimed at  identifying the extent of contamination,
544
STATE PROGRAMS

-------
rather man its existence, are beyond the scope of a site inspection.
If such activities are warranted, they are undertaken during site in-
spection follow-up activities.
  Under the RCRA 3012 program, 171 sites have been inspected
including 95 identified from the preliminary assessment stage. All
of these inspections involved  on-site  interviews with facility per-
sonnel, and many involved collection of soil,  sediment and  water
samples. Of these inspected sites, approximately 45% were deter-
mined to pose no  hazard to  the  environment  and therefore no
further action was recommended under the RCRA 3012 program.
  The remaining sites were assessed a low or medium hazard po-
tential, based on the type and  quantity of waste  materials present,
waste containment, pollutant migration pathways and potential re-
ceptors.  Many of the sites for which a medium hazard potential
was assigned were already under enforcement and no further action
was recommended under RCRA 3012.
  Relatively few sites, which had not already been investigated or
which were unknown to federal and state regulatory agencies, were
assigned a medium hazard potential.  However,  two sites  were
selected  for site  inspection  follow-up activities  due to suspected
problems.
Site Inspection Follow-Ups
  Inspection follow-up  activities are required where additional in-
formation is necessary to calculate or strengthen  a MRS score, bet-
ter define the quantity  of waste materials and the extent of con-
tamination or further identify on-site  waste materials. At many of
these sites, groundwater monitoring well installation is necessary to
provide information on subsurface transport of waste materials.
  Two sites were identified under the Texas RCRA 3012 program
for which additional information was desirable. The first site in-
volves an abandoned petroleum refinery which was dismantled in
the late 1940s. Several large impoundments were used for waste dis-
posal and/or crude storage during the active  life of the refinery.
These impoundments were filled, but much of the waste material
remains. Portions of the property have been sold, and a warehous-
ing complex was constructed over the largest of the former ponds
in the late 1970s. During construction,  quantities  of oily sludge
were uncovered, causing illness in several workers at the site and
causing  odor problems  in neighborhoods which  had developed
nearby. The follow-up inspection involves the collection of surface
soil samples to identify possible off-site transport of waste ma-
terials in surface drainage pathways. Soil borings and groundwater
monitoring wells are required to evaluate the subsurface geology
and potential subsurface transport mechanisms as well as the depth
and extent of buried wastes.
   The second site inspection follow-up involves an inactive landfill
formerly used  for the disposal of magnesium sludge from a non-
ferrous metal  alloy producer. At the time of disposal, this ma-
terial was still reactive and, in the past, had caused several fires in
the disposal area.  Although the inactive landfill is heavily over-
grown with scrub brush and a few small stands  of trees, extensive
piles of the magnesium  sludge are still evident.  On-site investiga-
tive activities will include soil and sediment sampling of drainage
pathways and  drilling of groundwater monitoring wells to  deter-
mine the potential for subsurface migration of heavy metals. Sam-
ples of the remaining waste material will also be collected to deter-
mine its current characteristics, including reactivity.

SUPERFUND ENFORCEMENT
  With the identification of hazardous  sites  under RCRA 3012,
there is a need for site restoration strategies. If sufficient documen-
tation exists to indicate which parties are involved, then an en-
forcement program is appropriate. In Texas, the TDWR has  devel-
oped such a program.
  Prior to the enactment of the Texas Solid Waste Disposal Act in
1969 and the RCRA in  1976,  the generation,  storage, transporta-
tion and disposal of hazardous waste was regulated only indirectly,
if at all, through a stretch-to-fit application of state water qual-
ity legislation and public health laws pertaining to nuisances. With
time, enforcement activity has increased in scope and complexity
and has expanded to include abandoned hazardous waste sites.
  The State of Texas gained the initial authority and resources to
respond to abandoned hazardous waste sites with the passage of
CERCLA in 1980. In 1981, the 67th Texas Legislature passed Sen-
ate Bill 758 establishing the Texas Disposal Facility Response Fund.
This fund provides the resources for the state to become a man-
aging partner with the federal government in the Superfund pro-
gram.  In February, 1982,  the Governor designated the TDWR as
the Agency with the necessary authority to develop and manage the
Superfund program in Texas.
  Based on its existing solid waste enforcement program and its re-
cent experience with  the federal  Superfund program,  Texas has
developed an enforcement policy which fosters early and direct
commitment, participation and site restoration by parties with doc-
umented involvement. Should such a procedure fail, the federal
Superfund resources are then tapped.
Enforcement Strategy
  The basic strategy in all enforcement cases is the use of enforce-
ment mechanisms of escalating significance. These mechanisms are
outlined briefly as follows. Timeframes and other details have been
omitted for brevity.
  Upon discovery, inspection and development of sufficient docu-
mentation, the TDWR District Office will send the affected party a
notice. This notice requires the recipient to come to an agreement
on  a proposed schedule  for  site  restoration. If an adequate re-
sponse is not received by the District Office, the case is referred to
the Solid Waste Enforcement Unit.
  The enforcement investigator reviews the case and inspects the
site as necessary. Upon coordination with TDWR General Coun-
sel, a conference is then held with the involved party in order to
develop a timely schedule  for site restoration. The mechanism for
attaining this goal is either a compliance agreement signed by both
parties or a letter signed by the Executive Director of TDWR. If an
agreement cannot be reached or violations of prior agreements are
not resolved, the case is referred to the Texas Attorney General or
to the TDWR Superfund unit. Restraining Orders to stop dumping
and referral to the USEPA for imposition of administrative fines
are two additional mechanisms which may be used.
Enforcement Work Scope

  TDWR currently has  approximately 3,000  Class  I  Industrial
Solid Waste generators and facilities, any one of which could be-
come involved in a Superfund site. As of June 30, 1984, the Solid
Waste Enforcement Unit  had a total worklist  of 321 cases, each
with a higher associated probability of becoming involved with
Superfund.
  A review of this worklist indicates that at least 20 (more likely 30)
sites are abandoned with a high likelihood of Superfund  involve-
ment. The 20 sites are in addition to the existing 11 NPL sites and
12 proposed NPL sites.
  Further examination reveals that enforcement action has pro-
duced results or is pending with involved parties in at least  11 of
these enforcement cases.
  The resolution of these cases has centered upon timely  commit-
ment and  participation by involved  parties with the option of
TDWR referral to Superfund. A number of the remaining aban-
doned site cases  do  not  have  a documented responsible party.
Apparently, this "Superfund inducement" is an effective enforce-
ment tool. The ultimate decision for the involved party is the ex-
penditure of a site restoration cost now versus the expenditure of a
probable higher site restoration cost under Superfund multiplied by
the probability of successful USEPA cost recovery action multi-
plied by a factor of up to three.
  Thus, the experience of TDWR suggests that a viable enforce-
ment program can use the "threat" of federal Superfund in a time-
ly and efficient manner to attain site restoration. This effectively
broadens the scope of the federal Superfund program at the state
enforcement level and indirectly  extends the federal program re-
sources.
                                                                                                 STATE PROGRAMS
                                                           545

-------
       HAZARDOUS WASTE  POLICIES AND  MANAGEMENT
                           PRACTICES  OF  NEW  YORK  CITY

                                                 CAREY WEISS
                                           STANLEY SIEBENBERG
                                               CHARLES SMITH
                          New York  City Department of Environmental Protection
                                            Office of Policy Analysis
                                              New York, New York
INTRODUCTION
  There has been a new role emerging for local governments in
remedying the problems posed by abandoned hazardous wastes.
As testimony to this emergence, one may point to the network of
new local regulations imposing penalties against midnight dumpers,
the recent activities of local government coalitions like a new organ-
ization called the National Association of Local Governments on
Hazardous Wastes (NALGOHW)  and the efforts that local gov-
ernments are making to equip themselves and to hire the technical
personnel needed to rapidly respond to abandoned waste incidents.
While Congress may have intended that federal and state govern-
ments take the lead in regulating and remedying hazardous waste
problems, the local governments now have reason to get involved
since the lack of State and Federal resources has left gaps which
must be filled.
URBAN DUMPING
  New York City has been quite active in this area since it is faced
with some very difficult and unique hazards. While there may be a
national perception that hazardous waste dumping occurs primarily
in isolated rural areas, there is an urban version which is just as
prevalent. This urban  problem is extremely difficult to resolve be-
cause  dumping occurs in small quantities and in numerous loca-
tions where the density of buildings is sufficient to provide cover
for illicit operations.
  What makes urban areas like New York unique is the population
and building density.  The City of New York is divided into 59
community districts, each district consisting of several neighbor-
hoods. To illustrate the magnitude of the City's population, each
of these 59 districts contains an average population  of 120,000
persons which means that one district alone is larger than most of
the country's 50 major central cities.  With an average of 540 per-
sons residing on each acre, hazardous waste problems which might
be considered insignificant  elsewhere  can pose serious health
threats; school children, vandals or other persons may come into
close contact with the substances. In addition, there are approx-
imately 1,000 businesses,  many of them small, which handle haz-
ardous chemicals; as a  result, hazardous wastes are present in small
quantities in many locations and in places which are  not readily
identified.

DISCOVERY
  Abandoned or uncontrolled hazardous substances are brought to
the attention of the City government in either of two ways. First,
                                                    citizens may call with eye witness information which may or may
                                                    not be reliable. Generally, the police are dispatched to verify these
                                                    calls.  In other situations, hazardous substances may be spotted by
                                                    the city's uniformed forces, including police,  fire, sanitation and
                                                    the medical services emergency units. Generally these uniformed
                                                    services personnel will discover a hazardous substance problem in
                                                    the course of their normal dutues. These calls are referred directly
                                                    to the New York City Department of Environmental Protection
                                                    (NYC DEP) for response.
                                                      In New York City, dumping occurs typically in the streets, in-
                                                    side buildings,  in vacant lots and secluded parkland areas and, in
                                                    the past, at the  municipal landfills.  Dumping within buildings
                                                    seems to take  place in  two  circumstances. The first is where a
                                                    building is used for dumping wastes which have been generated
                                                    elsewhere. This involves an illicit transporter in an unload-it-and-
                                                    run situation. Second, and quite prevalent, is the abandonment of
                                                    wastes after the bankruptcy or closing of a company. In this case,
                                                    a business generates wastes or uses hazardous materials at its loca-
                                                    tion for a number of years and then goes bankrupt, walking away
                                                    from both the property and the substances. The City of New York
                                                    has responded  to these and other related problems by developing
                                                    its own response, investigation, and litigation capabilities.
                                                    RESPONSE

                                                      Specific response activities include:
                                                    •A field team, recently expanded to include 17 chemists and other
                                                     technical staff, responds to calls about abandoned hazardous sub-
                                                     stances on a 24 hr day, 7 day week schedule. This team responded
                                                     to more than 225 incidents in  1983 and handled approximately
                                                     1000 calls from citizens and businesses seeking advice on haz-
                                                     ardous substance problems. In  addition, the team has seven lab-
                                                     oratory staff which  have  recently gone  on  a 16 hr per day
                                                     schedule.
                                                    »/4 Motor Carrier Safety Unit  enforces federal, state and local
                                                     hazardous substance transportation laws. From July 1983 to July
                                                     1984 this unit, which as 23 police personnel, issued over 5000 haz-
                                                     ardous substances related summonses.
                                                    •A Fire Department Hazardous Materials Unit is outfitted  for
                                                     hands-on response to chemical emergencies. This unit is equipped
                                                     with emergency items such  as overpack drums and encapsulated
                                                     suits. Although the city generally engages a licensed contractor
                                                     to clean up  abandoned wastes, the  Fire Department Unit pro-
                                                     vides immediate assistance while  arrangements are  being made
                                                     with the cleanup companies.
546
STATE PROGRAMS

-------
•A 15 member team of investigators and attorneys was recently
 created to conduct investigations into the identity of responsible
 parties in abandoned waste incidents.
•A team of four affirmative  litigation attorneys was also recently
 created to work on hazardous substance abandonment cases ini-
 tiated by the city.
  In total, the City of New York is spending $3-$4 million annually
on hazardous waste program  personnel  salaries and  almost $2
million annually on contracts for analytical work, disposal and re-
lated  cleanup, exclusive of testing programs underway at several
municipal landfills. This annual budget is perhaps double the en-
tire New York State Superfund budget.
 CASE HISTORIES

  Specific case studies will be described below to further illustrate
 the different types of waste abandonment problems which are con-
 fronted by the City of New York.
 Bankruptcies

  The city frequently becomes involved in cases where bankrupt
 businesses walk away from hazardous  wastes or materials which
 were generated while the business was active. Unfortunately, gov-
 ernmental entities are seemingly powerless to stop this type of waste
 abandonment since the bankruptcy laws, which are structured pri-
 marily to protect creditors, allow companies to walk  away from
 their environmental responsibilities.
 Quanta Resources

  Such was the case with the Quanta Resources site in Long Island
 City, which is in New York City's  borough of Queens. The State
 had entered into a consent order with this company, resulting from
 numerous environmental violations, but all attempts to bring the
 site into compliance had failed. On the evening of May 7, 1982, the
 NYC DEP received notification from the New York State Depart-
 ment of Environmental Conservation that the trustee for the bank-
 rupt  Quanta Resources Corporation  had  petitioned in federal
 bankruptcy court to allow an abandonment of  the property. The
 city was  further  notified that, upon granting of the petition, the
 trustee would remove  guard  security  and  fire prevention and
 other emergency equipment from the  Corporation's  oldest and
 most dilapidated waste oil reprocessing plant.
  Following the initial meeting and site characterization, both the
 State and Federal  governments ended their involvement with the
 site's remedial efforts.  The State  claimed that State Superfund
 monies,  which are available for inactive hazardous waste sites,
 could not be made available in this case as the site did not fit the
 precise definition of an inactive business as contained in the State's
 environmental law.
  On the Federal side, an initial USEPA evaluation using the Haz-
 ard Ranking System (MRS) assigned a  low priority to the site be-
 cause potable groundwater  supplies  were many miles away and
 neither population density nor poltential air contamination were
 weighted heavily by the HRS. In addition, there were many un-
 known factors, such as quantity and precise waste characteriza-
 tion which limited the  priority  ranking.  This,  coupled with the
 USEPA's policy of not taking action at sites where  other govern-
 mental entities are taking an active interest, sealed the city's fate.
  The city, therefore, was forced to secure the site and to perform
 preliminary analyses of the wastes  so that an assessment could be
 made of the seriousness of the hazard.  Abatement of this hazard
 ultimately involved the removal of over 640,000 gal of waste oils,
 sludges and water. Much of this material, which was stored in 106
 separate  tanks, was contaminated with  PCBs and other chemicals
 including corrosives and cyanide salts. The 6 month project  in-
 volved characterization of the initial waste stream, further site soil
 and groundwater characterization,  and  a risk analysis study along
 with clean-up and  disposal. The  total cost was $2.3 million, all of
 which was borne by the city.
  The city did,  however, press the federal  authorities for reim-
bursement, and  had initially submitted a request to USEPA for
emergency response assistance. When,  after  a 6 month delay, the
reply arrived, it denied the request,  citing a  failure to obtain
USEPA approval prior to initiation of the cleanup, which is a legal
requirement embodied in CERCLA. The city, however, had been
unable to wait for the USEPA response as low  flash points, four
story high tanks which were structurally unsound, 100°F  summer
weather, and a strong public outcry had mandated that at least pre-
liminary  action take place immediately. The city is still continuing
its efforts to identify sources of reimbursement, including an iden-
tification of the  companies which had  contracted for disposal of
their wastes using the Quanta Resources Corporation.
Berg Chemical

  In a second bankruptcy case, the city has been a little more suc-
cessful in gaining the promise of future reimbursement; this time
from a group of creditors, with the assistance of the Federal  bank-
ruptcy court. In the spring of 1984, NYC Fire Department brought
to the attention  of NYC DEP, the hazardous conditions  existing
at Berg Chemical Co., Inc., located in the South Bronx. The com-
pany was a chemical repackaging facility and a distributor of chem-
icals, food dyes and detergents; it had filed a petition for Chapter
11 recoganization under the federal bankruptcy law.  The  com-
pany had no valid permits as a hazardous waste generator, no dis-
charge permits,  and had been issued many  local fire violations.
Like Quanta Resources, Berg Chemical was operating pursuant to
a consent order  negotiated  with the State of New York  in July
1983.
  The city, fearing that the property would be  abandoned, con-
ducted  an inspection which revealed  many  potential dangers
created by poor housekeeping  practices.  For example, cyanide
compounds and  acids were  stored in the same area and food dyes
and additives were stored next to toxic  chemicals, many of which
were in damaged or leaking containers. In addition, numerous roof
leaks were apparent and there was no functioning fire prevention
or fire  suppression  equipment.  Finally, many  drums were un-
marked and company personnel did not know what was in most of
the unlabelled drums.
  In an effort to prevent abandonment of the property and over
12,000 containers  of various substances, the city applied to the
bankruptcy court for relief. In response, the court directed Berg to
segregate new products from waste materials; to properly store all
new  products; to clean up and dispose of all wastes; and to install
fire detection and suppression equipment. Unlike Quanta, where
the company had  already ceased  operations, Berg was an active
business and was permitted by the court to remain in operation.
  The bankruptcy court was then instrumental in getting the cred-
itors to agree to a plan where the city would go in and clean up the
site and  in return would be given superiority status to become
the first creditor to be reimbursed from the proceeds of the ultimate
sale of the land.
  The cleanup of this site was underway as of August, 1984 and is
expected to cost about $280,000.

Other Abandonments

  Abandoned buildings are owned by the city as a result  of non-
payment of property taxes. This property foreclosure process is
mandated by local law and is an  automatic  process initiated pri-
marily by a finance department computer. The action is triggered
after several years of non-payment of real estate taxes. Given the
large number of  these foreclosures and the automatic nature  of the
vesting process, physical site inspections do not generally take place
until after the city takes title to a property.
   Since the City of New York forecloses on hundreds of buildings
each year and currently holds an inventory of approximately 1200
commercial and industrial buildings, any type of hazardous waste
can be found. Where these wastes are  discovered, the city,  as  the
new owner, becomes involved in a cleanup which may cost tens, if
not hundreds, of thousands of dollars.
                                                                                                STATE PROGRAMS
                                                          547

-------
Technical Metals Finishing Corporation

  In one such case in  1982, the city was required to  prevent the
imminent mixing of acid with cyanide salts that were left in an
abandoned electroplating facility in Brooklyn which had been oper-
ated by the Technical  Metals Finishing Corporation. The imme-
diate removal of  12,000 gal of strong acids and base plating solu-
tions in open vats and  in damaged containers and the cleanup of
cyanide salts on the floor  was necessary due to the failure of the
company to take proper removal actions prior to abandoning the
property. The removal of these substances was deemed necessary
by the City to prevent  an immediate and significant risk to public
health in the  surrounding residential community and to the 300
children in the elementary school located to the rear of the electro-
plating facility. These actions initially cost  the city approximately
$56,000, but the funds were later recovered by the local district
attorney from the president of the corporation.
  Following these immediate response actions by the city, the site,
which  still contained contaminated  floor boards and earth, was
placed on the state's inactive hazardous waste site list for further
investigation and possible inclusion on the state Superfund list.
  A year later, when the city took title to the property for non-pay-
ment of back  real-estate taxes, the site had  not yet been evaluated
for Superfund status. As the new owner, the city was considered
responsible and was compelled to initiate a comprehensive site
cleanup and decontamination program which cost about $230,000.
While  the city  may attempt to gain reimbursement from the
previous owner or from the state Superfund, it is possible that the
city's ownership status will make reimbursement more difficult, if
not impossible.
Hospitals
  Abandoned hazardous materials are often found after a business
closes  or relocates. These materials, which are not generally regu-
lated by the same environmental statutes as  hazardous wastes,
effectively become wastes upon their abandonment. Such was the
case with many of the closed hospitals in the City of New York.
  In the period between 1974-1980, 46 hospitals were closed in the
city. Upon inspection, laboratories and pharmacies stocked with
useable chemicals were found in about one quarter of these  facil-
ities. This situation apparently  occurred because the state, which
has authority  over hospital facilities and which issues certificates
of operation, no longer has authority once a certificate is revoked
or expired. Thus,  the chemical supplies  were no longer useful
materials and had become wastes. No government agency,  how-
ever, had clear authority for  following-up in these cases to insure
proper cleanup and disposal of the wastes.
  This problem was initially  brought to the city's attention when
the Fire Department reported that biological specimens had  been
found  outside of the former Logan Hospital in  Harlem. Upon in-
spection, it was found that this abandoned hospital still contained
large amounts of chemicals, Pharmaceuticals, biological specimens
and compressed gases.  The city immediately placed security on the
buildings, and within several  days a certified contractor was hired
to clean up the site. Over 200 lab-packs along  with other poten-
tially hazardous  substances  were  removed from the site in the
course of the project.
  This incident initiated a year-long investigation into the other
closed  hospitals.  During this time,  NYC  DEP chemists visited
every room in all 46 former hospital  facilities. Where past or cur-
rent  owners could be identified, they were required to clean up any
chemical or biological wastes  which were found. Where no owners
could be identified or where the city had taken title to the property,
the city assumed the cost of cleanup. Total costs including security,
building  closure,  cleanup   and  disposal  were  approximately
$670,000.
                                                          Because of this experience, new procedures are now employed
                                                        whenever a hospital is about to cease operation. The state agency
                                                        which had originally issued the operating certificate calls upon the
                                                        city to conduct a joint  inspection of the property, and the certifi-
                                                        cate of operation is terminated only after all hazardous substances
                                                        have been removed.  In addition, the NYC DEP investigation en-
                                                        couraged the local Health and Hospitals Corporation to review its
                                                        storeroom inventories and to hire waste haulers to remove chem-
                                                        icals and pharmaceuticals no longer used. This was done to prevent
                                                        a build-up of "dead stock."
                                                        Other Types of Waste Incidents
                                                          There are numerous  other examples of hazardous wastes which
                                                        must be removed by the city including the following:
                                                        •Ethers, acids and cyanide salts were discovered as a  result of in-
                                                          vestigation of illegal drug operations; in the period from May to
                                                          July 1984, over 725 gal of liquid  chemicals and 70 Ib of solids
                                                          were collected by the city.
                                                        •Municipal landfills have been used  for illegal dumping in the past.
                                                          As a result, millions of dollars are now being spent by the city to
                                                          characterize air emissions and leachate from these landfills. Oper-
                                                          ating procedures have been significantly changed so that illegal
                                                          dumping is prevented.
                                                        •With access to the landfills stopped, dumping has increased on the
                                                          streets and roadways  and on parks and other vacant lands. The
                                                          city  has recovered  approximately  1500 containers of hazardous
                                                          substances since 1979 in 300 separate incidents of illegal dumping.
                                                        RECOMMENDATIONS
                                                          It is necessary for the State and Federal governments to join with
                                                        local governments like New York City in a partnership to solve
                                                        these urban hazardous waste problems. Specifically, the efforts of
                                                        local governments should be recognized and integrated into the na-
                                                        tional scheme by:
                                                        •Introducing into RCRA a "caretaker" status rather than a gen-
                                                         erator or responsible party status for local governments which
                                                         acquire  hazardous wastes through response to emergency  inci-
                                                         dents, property foreclosures or inadventent gifts of contaminated
                                                         land donated for parks or other municipal uses.
                                                        •Providing uniform protocols for all levels of government to use in
                                                         assessing the degree of contamination and the risks posed by un-
                                                         controlled hazardous waste sites, thus avoiding inconsistent tech-
                                                         niques and redundant efforts.
                                                        •Organizing federal or state technical assistance/technical trans-
                                                         fer mechanisms for localities to  use  when faced with hazardous
                                                         waste problems which do not  quality  for federal or state remedial
                                                         funding. Such assistance might include technical review of clean-
                                                         up contractor proposals, verification of proposed contractor qual-
                                                         ifications, consultation with affected communities and assistance
                                                         with identification of the responsible parties.
                                                        •Reviewing  the current site prioritization  procedures  for urban
                                                         area needs. Specifically, explosive or flammable situations where
                                                         the air  is a potential pathway  to citizens in densely populated
                                                         areas does not seem to be adequately addressed in the NCP, which
                                                         ispartofCERCLA.
                                                        •Amending the federal bankruptcy and environmental laws to in-
                                                         sure that the bankrupt company or  its successors in interest re-
                                                         main responsible for cleaning  up contaminated sites.
                                                          It  is essential  that  some of  these legislative and  regulatory
                                                        changes be examined. Local governments are on the front lines in
                                                        the war  against uncontrolled  hazardous wastes and can  offer a
                                                        great deal toward solving these national problems.
 548
STATE PROGRAMS

-------
      THE  NATO/CCMS  STUDY OF CONTAMINATED  LAND

                                                    M.A. SMITH
                                         Department of the Environment
                                         Building Research Establishment
                                       Garston, Watford, United Kingdom
 INTRODUCTION

  The NATO-CCMS  (Committee on Challenges of Modern So-
 ciety)  Pilot Study Group on  Contaminated Land adopted the
 following definition of contaminated land:
 "Land that contains substances that, when present in suffic-
 ient quantities or concentrations are likely to cause harm to
 man, the environment or on occasions to other targets."
  The emphasis on the presence of contaminants means that it em-
 braces the  uncontrolled hazardous waste sites of particular con-
 cern in the United States, many types of former industrial land
 (for example, metal mining,  chemical production and coal gas
 production) of particular concern in Western  Europe and  also
 land that has become contaminated due to  aerial deposition or
 through the application of sewage sludge "rich" in toxic metals.
  As previously described,1 the Study Group first reviewed the
 overall problem of contaminated land (identification,  assessment
 and remedial action) in order to decide where it could most  use-
 fully direct its efforts. It decided to concentrate on the identifi-
 cation of remedial measures that were effective in the long term.
 Accordingly, seven projects were established on different types of
 remedial action and on certain related topics.  Thus,  The Study
 was intended to examine  in detail  a number of important aspects
 of contaminated land rather than to cover the whole subject.  The
 Study Group has  now completed its work and its report2 will be
 published shortly. In this paper, the author  presents a brief ac-
 count  of the Study and some of its more important conclusions.
 Other papers in these proceedings describe the outcome of some of
 the individual projects.

 STRUCTURE OF THE STUDY

 The Participants

  Seven countries took part in the study (Table  1). Eleven CCMS
 Fellows contributed either directly to the main report or pro-
 duced  related reports  of their  own3' 4- 5 (CCMS fellowships are
 awarded annually for  work related to one of the current CCMS
 studies).

 The Projects

  The seven projects are  listed in  Table 2. The emphasis was on
 methods of dealing with contamination where it is found. The three
 main methods identified (on-site treatment, in situ treatment  and
 macroencapsulation) together with control  and treatment  of the
groundwater regime provided the  basis for four of the projects.
These four, on the practical  aspects, were primarily state-of-the-
art reviews within the context set by the project on long-term effec-
tiveness. The project on flammable and toxic gases also dealt with
remedial actions but additionally considered the nature of the haz-
ards and their identification and assessment. This latter aspect pro-
vided a link to the project on rapid methods of on-site analysis.
                          Table 1
  National Representative Members of the CCMS Pilot Study Group on
                     Contaminated Land

Canada, K.A. Childs, Senior Adviser, Landfill Site Remediation, Environ-
  ment Canada
Denmark, Ms. K. Warnoe, Environmental Protection Agency
France, P. Godin, Direction de la Prevention des Pollutions, Ministere
  de L'Environment
Federal Republic of Germany, K. Stief, Umweltbundesamt
Netherlands, J. van Lidth de Jeude, Ministry of Public Housing, Physical
  Planning and the Environment
United Kingdom, M.J. Beckett, Central Directorate on  Environmental
  Pollution, Department of the Environment
United States of  America,  D.E.  Sanning, Municipal Environment  Re-
  search Laboratory, United States Environmental Protection Agency
  The final report also includes a chapter based on a study by one
of the CCMS Fellows on the problems of redeveloping old iron and
steelmaking sites in order to illustrate the value of such industry/
contamination audits in giving warning of potential problems. The
steel industry was chosen because of the large areas of land becom-
ing  derelict as the industry contracts in both North America and
Western Europe.
Method of Working

  The participant countries  chose to lead projects in which they
already had an interest so that the work done would be of benefit to
their national programs. Draft reports were reviewed by corres-
pondence and at meetings of the Study Group. Finally, the indi-
vidual project reports were brought together by the Study Director
as Chapters in the final report2 which will be published by Plenum
Publishing Corporation in 1985. The eleven technical chapters de-
rived from the project reports are supported by an  introductory
chapter and by a chapter summarizing the conclusions and recom-
mendations of the Study Group. The latter provided the basis for
the following sections of this  paper.
                                                                                 INTERNATIONAL ACTIVITIES
                                                       549

-------
                             Table 2
     List of Projects Carried Out as Part of the CCMS Pilot Study on
                       Contaminated Land

A  In situ Treatment of Contaminated Sites
  Methods of treating the bulk material on a contaminated site without excavation
by detoxifying, neutralizing, degrading, immobilizing or otherwise rendering harm-
less contaminants where they are found.
Project Leader:
  D.E. Sanning
  USEPA

B  On-site Processing of Contaminated Soil
  Methods of decontaminating or otherwise reducing  the potential environmental
impact of the bulk of contaminated material on a site by: excavation; treatment to
detoxify, neutralize, stabilize or fixate; and, usually, redeposition on-site.
Project Leader:
  J.W. Assink
  TNO, Netherlands

C  Cover and Barrier Systems
  Systems designed to prevent the migration of contaminants vertically or laterally or
to prevent ingress of surface or groundwater into contaminated sites.
Project Leaders:
  (a) Covering systems:
       Dr. G.D.R. Parry
       Environmental Advisory Unit
       Liverpool University, UK
  (b) In-ground barriers:
       K.A. Childs
       Environment Canada

D Control and Treatment of Groundwater
  Primarily concerned with those operations designed to control or treat the liquid
phase on contaminated sites including design  of cut-off systems, hydrogeological
modelling and groundwater treatment.
Project Leader:
  K.A. ChUds
  Environment Canada

E  Rapid On-site Methods of Chemical Analysis
  Methods of chemical analysis allow determinations to be made on "soil", water
and air samples on-site to speed and reduce the costs of site investigation.
Project Leader:
  M. Gruenfeld
  USEPA

F  Long Term Effectiveness of Remedial Measures
  Overall problem of the design of long term effective remedial measures. Collection
of information on examples of remedial and restoration actions that have demon-
strably worked for a number of years and methods for the evaluation of sites for long
term effectiveness of remedial measures.
Project Leader:
  K. Stief
  Umweltbundesamt, FRG

G Toxic and Flammable Gases
  Concerned with:  volatile organic emissions from contaminated sites, production,
migration and control of gases from land disposal sites including typical landfill gases
such as methane and carbon dioxide.
Project Leader:
  S. James
  USEPA	

GENERAL CONCLUSIONS

   The  Study Group started from the premise that excavation  and
removal of the contaminated material for deposition elsewhere  is
not always environmentally acceptable or practicable. In addition,
it may not be a "once and for all" solution as the disposal site may
itself become a "problem" in the future.'
   Remedial actions fall into three main groups:
•Those that remove contaminants or render them harmless
•Those that prevent the release of contaminants
•Those that reduce the rate of release of contaminants
   Those falling in the first group are preferred. On-site process-
ing and certain forms of in situ treatment may provide such solu-
tions; the further development  of these technologies should be en-
couraged.  Provided  these technologies are applied properly, their
long-term effectiveness is secured.
   The  second  process is not achievable  in  practice, although in
situ treatments in which the contaminant is chemically converted to
an insoluble form under  forseeable environmental conditions may
                                                           provide an essentially permanent solution provided the difficulties
                                                           of application of in situ techniques can be overcome.
                                                             In practice, therefore, all technologies, other than those in which
                                                           the contamination is destroyed or rendered harmless, offer a solu-
                                                           tion of only limited or uncertain duration unless other mechanisms,
                                                           such as microbial attack, reduce the contamination. In general,
                                                           treatment systems based on isolation (e.g., covering systems and in-
                                                           ground barriers) are likely to lose effectiveness with time and, like
                                                           most other civil and structural engineering works, have a finite life.
                                                           They will need monitoring, maintenance and renewal as long as the
                                                           contaminants are present and their  release would be considered
                                                           harmful. This is  analogous to the  monitoring of many civil engi-
                                                           neering construction projects (e.g., bridges and dams) and should
                                                           not be viewed as  casting doubt on the effectiveness of the selected
                                                           solution. Most structures also receive regular maintenance. Fund-
                                                           ing arrangements should  take into account the need  for monitor-
                                                           ing, maintenance and  the need sometimes for a phased approach
                                                           to remedial action.
                                                             The term "effectiveness"  can be used to mean different things
                                                           during the stages  of any remedial measure. It can be applied to the
                                                           performance of a component part of the remedial system (e.g., cut-
                                                           off wall) or to the system as a whole (e.g., cut-off wall plus ground-
                                                           water pumping); a distinction can also be made between theoretical
                                                           effectiveness and installed effectiveness; and  long-term effective-
                                                           ness can be assessed on an arbitrary  scale at a point in time or as
                                                           the ability of the system to continue to perform to an acceptable
                                                           standard over a  prolonged period of time ("performance"  is a
                                                           better term for this latter concept).
                                                             Very few of the technologies reviewed have been sufficiently
                                                           proven in applications specific to the treatment of contaminated
                                                           land, although they may  already be  in use for other purposes. It
                                                           is essential, therefore, to  carry out  proper long term evaluation
                                                           studies.
                                                             New methods of treatment cannot be proved satisfactorily in the
                                                           laboratory alone. There is a  need for properly designed and eval-
                                                           uated field  trials and  demonstration projects. Such projects are
                                                           more likely to be accepted if they are underwritten by responsible
                                                           authorities.
                                                             The evaluation of remedial actions requires the establishment of
                                                           an adequate set of data before and immediately after treatment and
                                                           the monitoring of parameters that will:
                                                           •Describe the behavior of the remedial system
                                                           •Describe conditions within and outside the contaminated area
                                                             The Study Group drew a sharp distinction between monitoring
                                                           and evaluation: monitoring is concerned with whether the remedial
                                                           system is working properly and is best regarded as  a component
                                                           part of the system; evaluation is closely related  to research and may
                                                           require elaborate instrumentation and inspection.
                                                             Retrospective studies of already reclaimed sites may provide  use-
                                                           ful information  on the performance of the treatment strategies
                                                           adopted but may be difficult owing to a lack of baseline data and
                                                           an unwillingness by the responsible authorities to have any doubts
                                                           cast upon the "success" of a completed reclamation scheme.
                                                             The long term nature  of  research into remedial  measures de-
                                                           signed for containment or stabilization must be recognized. Some
                                                           projects may need several  years for completion, and continued ob-
                                                           servations may be required over decades; research funding should
                                                           reflect this time scale.
                                                             Records of contaminated sites and of the treatments carried out
                                                           should be kept unless the  contaminants are removed or destroyed.
                                                           Such records are important since the use of the site may change,
                                                           the remedial measures deteriorate or knowledge about the effects
                                                           of the contaminants may change, thereby changing  one's  percep-
                                                           tion of the risks.

                                                           RECOMMENDATIONS  TO CCMS

                                                             The  Study Group made  a number of recommendations1 to
                                                           CCMS; the chief recommendation was that CCMS should encour-
                                                           age member governments  to consider the adoption of policies that
                                                           will:
550
INTERNATIONAL ACTIVITIES

-------
•Minimize the occurrence of contaminated land problems in the
 future
•Abate the adverse environmental impacts from contaminated land
•Allow for the safe reuse of contaminated land
  The first policy is intended to include avoiding the creation of
wastes; taking long term after-use of disposal sites into account
from the outset, and giving due consideration to the location and
operation of industrial plants to avoid contamination. The second
policy covers the need to carry out remedial actions on sites once
identified. The third covers a need to identify contaminated sites
and to control their use if the contaminants are not removed or ren-
dered harmless.

INDIVIDUAL PROJECTS

  The sections below are intended to highlight  some of the im-
portant conclusions and more promising prospects.

Long Term Effectiveness

  This Project provided the basis of the discussion above. In addi-
tion to a  discussion of the philosophical and technical aspects of
long term effectiveness of remedial measures in general, a detailed
assessment was made of each of the remedial measures discussed in
other parts of the report in  terms of the opportunities they pro-
vide for permanent or long term solutions. A review was also made
of the methods available for assessing the performance of remedial
measures.

 On-site Processes

   In on-site treatment, the contaminated soil is excavated, cleaned
 up in some way and then re-deposited. In principle, there are sev-
 eral different  procedures which might be used: extraction, thermal
 treatment, chemical treatment, mechanical and physical separa-
 tion, steam-stripping, microbiological treatment, stabilization and
 flotation.
   On-site treatment processes, with the exception of those involv-
 ing stabilization, are designed to provide a  final solution. In gen-
 eral, they will rely on  the  application of established technology
 from the fields of chemical engineering, hazardous waste treatment
 and mineral processing. The prospects from a technical viewpoint
 are promising. Some processes involving thermal treatment and
 separation have already been applied successfully, and others are at
 an advanced  stage  of development. In the medium term, micro-
 biological treatment systems look promising.
   The Dutch government  has chosen  on-site treatment and the
 comparable concept of soil treatment at a central processing plant
 as the best long term option for dealing with contaminated sites in
 the Netherlands. It is encouraging the development of the neces-
 sary technology by direct funding of research and development and
 by funding of reclamation projects, some  of which are used to
 demonstrate and evaluate new methods and processes.8

 In Situ Treatment

   In situ treatment of contaminated  land  in which the contam-
 inated ground is treated without  excavation offers a number of
 attractive options for: (1) removal or destruction of contaminants,
 (2) stabilization  of the contamination  and  (3) solidification to
 achieve some engineering objective such as improved ground stabil-
 ity. Two main treatment methods are possible:
 •Surface application of treatment agent
 •In-ground injection of treatment agent
   The latter is analogous to the well-established  engineering  prac-
 tice of grouting which is one option.
   The major  technical difficulties are: (1) how to ensure intimate
contact between treatment agent and contamination which is com-
pounded  by the inherent chemical and physical heterogeneity of
many contaminated sites,  (2)  possible unwanted interactions be-
tween treatment agents and contaminants, (3) difficulties in ensur-
ing that treatment has been fully effective, (4) difficulties in apply-
ing injection techniques at depths of less than about 2 m and (5)
production in many cases of a liquid waste stream requiring treat-
ment.
  There have been few successful applications of in situ treatments,
and there are significant difficulties to be overcome. Nevertheless,
some interesting and promising developments are taking place. As
in situ  treatment concepts can, in many cases, offer the possibility
of permanent solutions, it would be worthwhile persevering with
research  and development in this area. Microbiological  treatment
systems seem to have considerable potential. Electro-osmotic tech-
niques and thermal treatment by electrical heating also merit
further investigation.
Barrier Systems and Hydraulic Measures

  At present, in most countries, attempts to solve  a "contami-
nant" problem are likely to involve cover, barriers and hydraulic
measures with provisions for treatment of groundwater and leach-
ate. Such measures may, in any case, be required to supplement on-
site and in situ treatment processes. Long term effectiveness is of
paramount importance for such remedial measures. The designer
has to consider the likely installed effectiveness of each component
of the scheme, its interaction with other components, the effec-
tiveness of the scheme overall and any changes that may occur with
time. This concept requires a systematic analysis with an identifica-
tion of risks (to the system) and their quantification where pos-
sible (e.g., 1 in 100 years rainfall events are commonly taken into
account in the design of drainage/sewerage schemes). This process
is analogous to any major engineering  design project. The design
analysis may be aided by the use of modelling of groundwater and
contaminant movement including the  effects  of barrier and hy-
draulic systems.
  The performance of covering systems is particularly time-depen-
dent. The component parts of the system and contamination may
change with time, and the environmental stress on the system may
also increase. Vegetation growth can be both beneficial and detri-
mental; for example,  increased  cover will reduce erosion but root
growth may lead to penetration of synthetic barriers and to uptake
of toxic elements. Synthetic materials are likely to deteriorate with
time.
  Vertical barriers to control the movement of groundwater and
contamination can be installed using well-established engineering
procedures including slurry trenches, diaphragm piling and grout
curtains. Such barriers will always permit passage of some waste or
other fluids,  either because  the permeability, although very low,
is nevertheless finite or because of unavoidable imperfections in in-
stallation. There are, however, doubts concerning vertical barriers'
long term effectiveness  as a means of controlling contamination
owing to possible adverse interaction of barrier materials with con-
taminants and the possibility of breaches either induced by nature
(e.g., tree roots) or by man (e.g., subsequent excavation). Horizon-
tal barriers can be installed by  means  of a number of ground in-
jection/grouting techniques, but these  techniques are not well-es-
tablished or proven. These  also are susceptible to adverse inter-
actions with contaminants.
  While there is a shortage of information on the long term effec-
tiveness of containment (macro-encapsulation) systems, the pros-
pects are improving. The construction of vertical barriers is a well-
established engineering  technique, the potential limitations with
interactions with regard to contaminants have been  recognized
and work is in progress to produce improved systems. Similarly,
the multi-functional nature of most covering systems has now been
recognized and  their design is progressively on a more rational
basis.9 Thus, while containment with associated hydraulic measures
may not provide a permanent solution,  it can often provide a solu-
tion that is likely to remain  effective for a considerable period of
time. During the "breathing space" thus provided, the hazard pre-
sented by the contaminants may be reduced by natural processes
and new forms of permanent treatment may be developed.
                                                                                     INTERNATIONAL ACTIVITIES
                                                          551

-------
Groundwater Management and Treatment

  In this project, the investigators examined conditions where
groundwater may be, or has been, adversely influenced by migrat-
ing contaminants and  the methods available to:  (1) enhance the
quality of the degraded groundwater by in situ treatment, (2) mod-
ify groundwater regimes and (3) treat the groundwater after extrac-
tion.
  A wide range of methods is available for the treatment  of con-
taminated water after extraction, but economic and technical via-
bility is  restricted by increasing  difficulty  in  treating even more
dilute contaminants. After treatment, the extracted water can be re-
injected  or discharged to surface  waters.  In  situ treatment of
groundwater suffers many of the same constraints as in situ treat-
ment of ground including difficulty in contacting reactive agents
and contaminants,  uncertainty  about permanence of treatment
and difficulty in  establishing effectiveness  while it is carried out.
Nevertheless, it has  been successfully employed on a  number of
occasions, and such techniques should be further developed.

Toxic and Flammable Gases
  Volatile organic compounds (VOCs) can  enter the environment
from many different sources by a variety of routes. Both  sources
and routes  may  differ at  different stages  of investigation,  and
remedial action and  the impact may occur  at significant distances
(up to several kilometres) from a site. Thus, it may be difficult and
expensive to identify all environmental impacts and, conversely, it
may be difficult to link an overt health or environmental impact to
a source.
  The potential long term  environmental  impact of  VOCs  and
ways of controlling their emissions  from sites requires further in-
vestigation. Current  dispersion models do not always give  reliable
predictions of contaminant movement.
Rapid Methods of On-site Analysis

  In the case of VOCs and other gaseous contaminants, field meas-
urements are essential. The project on  rapid methods of on-site
chemical analysis  dealt only with U.S. experience. Information was
sought from other countries, but little was forthcoming  reflect-
ing the different scale and nature of the problems in the different
countries.

Iron and Steel Making Sites

  The social consequences that result from a large-scale run-down
of an industry such as the iron and steel industry make it imperative
that the land is brought back into beneficial use. The elimination
of an immediate environmental impact is not enough; the land has
to be restored in  a way  which requires only minimal  long term
attention.
  The steel industry was selected because of the rapid reduction in
its size owing to improved productivity,  the effects of the present
world recession and increased pressure from imports from new pro-
ducer countries. In addition, much information relevant to coking
plants is already available in the audits made of the problems pre-
sented by coal  carbonization and similar sites.'" The size of the
potential problem can be seen in the 1980  contrast between pro-
duction and capacity  in the EEC countries: 128 vs 202 M  tonnes.
  Generalizations about iron and  steel making operations are diff-
icult because of the many technical  changes that have occurred in
the industry. Location, age and integral complexity all  have to be
considered.  However, they all have  in common, albeit to varying
degrees, a number of potential problems in terms of contamination
or engineering factors affecting redevelopment. Their importance
depends upon a number of site-specific factors including intended
site use.
  The principal problems include chemical  contamination  in the
coking plant area, pickling plant, ore and waste disposal areas such
as lagoons and slag  heaps,  large deposits  of old  physically and
                                                        chemically unstable slag, massive foundations and the presence of
                                                        underground workings and mineshafts when plants are located in
                                                        mining areas.
                                                        CONCLUSIONS
                                                          A study such as this cannot, by itself, produce new solutions to
                                                        the technical problems presented by contamination. Nevertheless,
                                                        the Study Group believes that its report will help in four ways by:
                                                        •Promoting an awareness of the need to consider the long term
                                                         effectiveness of remedial measures and by providing a philosoph-
                                                         ical and technical framework within which they may be judged
                                                        •Illuminating some of the interactions that  must be taken into
                                                         account
                                                        •Drawing attention to progress in developing permanent and long
                                                         term effective remedial measures
                                                        •Encouraging continued information and technology exchange be-
                                                         tween countries, organizations and technical experts
                                                          The assessment and treatment of contaminated land are quickly
                                                        developing subject areas of increasing interest in each of the par-
                                                        ticipant countries. Thus, during the course of the study, a number
                                                        of guidance manuals or handbooks have been produced and na-
                                                        tional conferences held. In general, the handbooks, like the CCMS
                                                        study, are based on desk reviews of the existing technology with
                                                        only limited  input from practical use. If the Study Group's recom-
                                                        mendation concerning the need for detailed  long  term evaluation
                                                        of remedial  measures  is accepted by  responsible authorities, then
                                                        the technical basis of such guidance documents should improve in
                                                        future years.
                                                        ACKNOWLEDGEMENTS
                                                          This paper has been prepared as part of the research program of
                                                        the Building Research Establishment and is published by permis-
                                                        sion of the Director. The author would like to thank the national
                                                        representatives, the project  leaders and the other members of the
                                                        Study Group for their contributions to the  report on which the
                                                        paper is based. Thanks are also given  to the national authorities
                                                        and to CCMS for providing the resources for the study.

                                                        REFERENCES
                                                         1. Smith, M.A. and Beckett,  M.J.,  "An International Study of Con-
                                                           taminated Land", Proc. of National Conference on the Management
                                                           of Uncontrolled Hazardous Waste Sites, Washington, DC, Nov. 1982,
                                                           431-433.
                                                         2. Smith, M.A. (ed.). Contaminated Land, Plenum Press, New York and
                                                           London (to be published).
                                                         3. Coldewey, W.G. (ed.), Unlersuchungen zur Wasserdurlass-bindiger
                                                           Boden (Investigation of the  water permeability of cohesive soils),
                                                           Mitteilungen der Wesifaelischen  Berggewerkschaftskasse, No. 43,
                                                            1983.
                                                         4. Schoettler, U., Behandlung von Kontaminiertem Grundwasser bei der
                                                           Saniemng von Altlasten (Treatment  of contaminated groundwater
                                                           from contaminated sites), 1984.
                                                         5. Coldewey, W.G., Experiences with the covering of contaminated land
                                                           in the Ruhr region, Westfaelische Berggewerkschaftskasse, 1984.
                                                         6. Barry, D., Atkins, W.S. and Partners, Epsom, UK (to be published).
                                                         7. Bernard,  H., "Love Canal  2030 AD", Proc. of National Conference
                                                           on the Management of Uncontrolled Hazardous Waste Sites, Wash-
                                                           ington, DC, Nov. 1980, 22-223.
                                                         8. "N Grondige  Aanpak" (A thorough approach-what you must know
                                                           about soil reconstruction), Ministry of Housing and the Environment
                                                           (Netherlands), The Hague, 1983.
                                                         9. Cairney,  T.C., "A  Rational  Approach to the Design of Cover  for
                                                           Contaminated Sites", Proc. Conference  Contamination of the En-
                                                            vironment, London; CEP  Consultants Ltd., Edinburgh,  1984, 294-
                                                           299.
                                                        10.  Wilson, D. and Stevens, C., Problems arising from the Redevelop-
                                                           ment  of Gasworks and Similar Sites,  UKAEA Harwell Laboratory
                                                            Report AERE-R-10366, HMSO, London, 1982.
552
INTERNATIONAL ACTIVITIES

-------
               DESECRATION AND RESTORATION  OF THE
                                LOWER SWANSEA VALLEY

                                             E.M. BRIDGES, Ph.D.
                                         University College of Swansea
                                   Swansea, South Wales, United Kingdom
INTRODUCTION

  The 480 ha of industrial dereliction in the Lower Swansea Valley
resulted from almost 250 years of smelting and processing metals.
At different times, copper, lead,  silver, arsenic and  zinc were
smelted in 22 plants along the tidal reach of the River Tawe. Later,
ten steel and tinplate operations were  established further inland
(Fig.  1). Up to 1924, the non-ferrous metals were most important
with two-thirds of the copper imported into Britain being smelted
in this one area.  Copper smelting  declined after 1880 and was
replaced  by zinc. After 1928, steel  and  tinplate assumed a
dominating position in the industry of the valley with four out of
every five British tinplate workers employed within twenty miles of
Swansea.  The legacy  of dereliction  included  virtually all  the
characteristic features such as the 7,000,000 tons of slag, ruined
buildings, restricted access and unvegetated, contaminated, eroded
soils of the valley side (Figs. 2 and 3).
  The Lower Swansea Valley Project was set up in 1961 with the
financial assistance of  the  Nuffield  Trust,  Swansea  County
Borough Council,  the Welsh Office and the University College of
Swansea.  The brief of  the  project was  "...to investigate  the
physical, social and economic situation  in  the Lower Swansea
Valley, to understand the reasons which had inhibited its develop-
ment in the past and to provide the information necessary for its
future development." The project was clearly seen as the first stage
in which information was to be gathered and interpreted, leading
eventually to the renewal of the devastated land and development
of new forms of land use. The work of the Lower Swansea Valley
Project is summarized in a final report.'


INVESTIGATION OF THE
PHYSICAL ENVIRONMENT
  Existing plans of the Ordnance Survey had insufficient detail for
the base work of the project. Aerial photography of the valley was
flown in  June 1962 and a detailed map compiled by photogram-
merty at a scale of 1:5000. A contour interval of 10 ft up to 200 ft
OD was adopted with a 20 ft interval above that level. So that the
position of former smelting works could be accurately located, in-
formation available on the existing 1:1250 Ordnance Survey plans
was superimposed on to the photogrammetric map. The availability
of a correct detailed base map enabled calculations to be made of
the area and volume of the many tip complexes. To assist this work
and  the subsequent  geological,  ecological,  pedological  and
hydrological contributions, the project area  was  subdivided  into
plots where common problems occurred. These plots have subse-
quently formed the basis of the reclamation plans.
                      y Upper Forest and Worcester
                    t \I1848-1958)  Ab.r „„„Ł,_,
                    ija	•
                          Dyffn/n(1874-1961)
>          Landore
        (1717-18761*
        Little Landore

                "»"Landore Siemens(l869-18881
Morfall 835-1 9241*


 Hafod(1B10-1980]
                                    Swansea Vale
                         .1860-1946,  ^876-1974)

                                 Q Vilhers(1873-c1929)

                                D Glamorgan!'-1907)
                      Q Glamorgan/Dillwvn(1720-1926)

                      <&    '
                      § Llansamlet(cl866-1905|

               indore|1869-1980)
                                   1km
                                 , 757 _,930)
                       Middle Bank(1755-1924|

                       White Rock(1737-1928)
                               Steel and/or Tinplate
                               Copper
                               Spelter (Zinc)
                               Other Non-ferrous
                               Railways \
                               Canal
                               in 1930
                         Figure 1
         Metalliferous Works in the Lower Swansea Valley

  In the early 1960s it was thought too expensive to remove the
many slag heaps, so experimental work began to find how plants
might be encouraged to grow in the inhospitable environment of
metalliferous wastes. If a satisfactory cover of plants could mantle
the tips, this would be a relatively inexpensive way of improving the
visual appearance of the valley.2 The use of amendments such as
sewage sludge, domestic refuse, pulverised fuel ash and inorganic
fertilizers was investigated in a series of field and plot experiments
with mustard, common bent grass and a grass ley mixture. This
work was essentially empirical  as little was known at that time of
the physico-chemical factors which controlled the availability and
                                                                                 INTERNATIONAL ACTIVITIES
                                                        553

-------
                                Active Indultry 1962

                                Active Tipping

                                Tipj

                                Raifwavl «nd Railway land
                           Figure 2
          Metalliferous Wastes in the Lower Swansea Valley
uptake of heavy metals by plants growing in toxic tip material. The
best results were achieved on the steel wastes and poorest growth
occurred  on zinc waste  with  copper showing an  intermediate
response.'
  Trials of different species of shrub were undertaken to ascertain
whether their growth would be satisfactory, with appropriate en-
couragement, to cover tips. Plants which were known to be tolerant
of acid conditions, atmospheric pollution, drought and exposure
were chosen. Some of these plants are natural colonizers of derelict
areas,  and  some  are nitrogen-fixers. They  include  Ligustrum
vulgare, Buddleja davidii,  Hippophae rhamnoides. Rhododendron
ponticum, Salix repens,   Clematis  vitalba,  Lupnius arboreus,
Robinia pseudoacacia, Sorbus aucuparia, Thelycrania sanguinea,
A In us  glutinosa,  Beta la verrucosa,  Medicago (innoculated with
Rhizobium sp) and Tagetes minuta and were grown in experimental
areas on copper, steel and zine waste tips.
  In parallel with this work, plant species were collected from soils
known to be rich in  copper,  lead  or zinc in which a  natural
tolerance to heavy metal levels higher than normal had evolved. By
cultivation and seed collection  it was hoped to obtain  tolerant
clones  of plants in sufficient  amounts  to use in revegetation.
Although  this aspect of the work showed promise, it was not pos-
sible to follow it through  during the  period of the project; subse-
quently the National Seed  Development Organization did  produce
the lead-zinc tolerant variety ofFestuca rubra known as "merlin".
  The  microbiology of the three main  tip types in the Lower
Swansea Valley was investigated using  the microbial colonization
of cellulose film and dilution  plates with a  variety of  cultural
media. Bacteria of several different groups were isolated, but the
nitrogen-fixing Azotobacter was not detected. Mycorrhizal  fungi
were absent from the surface materials of the tips. Investitations of
                                                                   the  soil respiration suggested that the  inhibition of soil-living
                                                                   microorganisms was more  through lack of organic matter than
                                                                   metal toxicity.
                                                                     A survey of plant species growing naturally upon the various tip
                                                                   materials in the Lower Swansea Valley was made. The percentage
                                                                   of ground covered and the number of species occurring in quadrats
                                                                   on the surface of the tips gave the following results:
                                                                                             Zinc Waste
                                                                                                Copper
                                                                                                 Waste
Steel Watte
                                                         "to of ground covered
                                                          including mosses                10         27          16
                                                         1* of ground covered
                                                          excluding mosses                 1           4          3
                                                         Total no. species
                                                          recorded                       4          15          36
                                                         No. of quadrats                 SO         57          70
                                                         Ave. no. per quadrat            0.04         1.5         3.2
                                                         (After Weilon ««/.')

                                                          A sparse cover of common bent grass (Agrostis tenuis) and wavy
                                                         hair grass (Deschampsia flexuosa) occurred  upon  the  exposed,
                                                         stony subsoil material of the eroded soils surrounding the smelters'
                                                         (Fig. 3).
                                                          A major emphasis originally in the Lower Swansea Valley Pro-
                                                         ject was to improve the physical appearance of the environment by
                                                         tree  planting in these eroded natural soils. After liming and some
                                                         fertilizing to encourage rooting,  Japanese larch, Lodgepole pine
                                                         and birch were planted from 1963 onwards. In the decade 1963-73,
                                                         over 100,000  trees on 37 ha were planted at minimal cost with
                                                         volunteer labor and trees provided by finance from the City Coun-
                                                         cil and the Welsh Office. These trees now make a major contribu-
                                                         tion  to the improved appearance of the valley. (Certain areas were
                                                                                  Figure 3
                                                                Soil and Gully Erosion in the Lower Swansea Valley
554
INTERNATIONAL ACTIVITIES

-------
planted with grasses but with less success.) Since 1973, emphasis
has been on landscaping funded by grants  from the European
Regional Development Fund as part of the infrastructure develop-
ment, with the major restoration being paid for by  the Welsh
Development Agency from government funds.
SOCIAL ASPECTS

  Investigations of the human environment  played  a  significant
role in the initial research. A transportation study found that the
valley floor was almost totally devoid of roads with large areas only
accessible  on  foot. The  River  Tawe formed a barrier along the
western  side of the project area, and the presence of railways and
disused canals broke the land into a number of small awkwardly
shaped areas (Fig. 2). The recommendations  for a new road net-
work for the valley included a spine road to serve all the northern
part of the project area, a new crossing of the River Tawe in the
Hafod-Pentre Chwyth region and improvements to the roads  on
both sides of the valley. The route of the western road improve-
ment should follow the line of the Swansea Canal to bypass Mor-
riston.
   The social survey  of the valley and surrounding districts exam-
ined the distribution  of population according to sex, marital status,
occupational status, educational  attainment,  income, children in
receipt of  free school meals, children taken into  care and juvenile
delinquency. The housing stock was found to be old and in need of
 replacement or renewal.  Schools  and other public buildings were
inferior to those found elsewhere in the town. There was a shortage
of public open space, and it was recommended that this should be
taken into consideration in the preparation  of  the  development
plan. Some improvements of the roads took  place, but otherwise
 no obvious results followed from  the social surveys. However, one
 of the most significant features of the development plans drawn up
by the Swansea City Planning Department is the "park" concept
which goes a long way to meeting the earlier  proposals for public
 open space in the valley. One such park is the Leisure Park occupy-
 ing land south of the main railway line with facilities for informal
 activity and organized sport.
                          Figure 4
   Land Use Proposals in the Lower Swansea Valley Project Report
  The Lower Swansea Valley Project was unique in many ways as it
was the first thorough investigation in Britain of the reasons for
dereliction and its persistence. The project brought together people
from many disciplines who contributed skills to unravel the history
of dereliction and to establish a wealth of data on the physical en-
vironment and human attitudes to it. The assembly of this material
enabled a draft plan to be suggested for the redevelopment of the
valley (Fig. 4).
  The opportunity to assess the success of the Lower Swansea
Valley Project came in 1979, when a conference was held to review
progress and to encourage the completion of the task of rehabilita-
tion of the valley.5 Although weighted strongly in favor of planning
and social studies, this conference indicated that research work into.
the environmental background was continuing but at a lower inten-
sity.6'7 Swansea City Council had commissioned studies on  the
River Tawe, the extension of Pluck Lake as an amenity feature and
on basic development plans. Further aerial surveys were flown in
1969 and 1982.  With the  announcement of the Enterprise Zone in
1980, investigations  of bearing  capacity of certain sites  for in-
dustrial development have  been undertaken specifically with  the
aim of providing information for prospective site occupiers.
  In retrospect, the most successful part of the Project has been the
tree planting which has transformed the appearance of the valley
over the past twenty years.  The experimental work to try to grow
plants on inhospitable tips was of academic interest, but the find-
ings were not to solve the problem. Investigative work for the pro-
ject gave an opportunity for training to several of the  leading
figures in land restoration in Britain today.

RECLAMATION
  Plans for Swansea after the Second World War had envisaged a
predominantly industrial  future for the valley, at a time when  the
major industrial concerns were  still active in  the valley.  Events
overtook this plan of 1960, and the findings of the project were in-
corporated  in a 1968 Draft Development Plan which made im-
provement  of the River  Tawe  a major feature. This, too,  was
changed by a 1974 consultants' report which recommended a mix-
ture of industrial and leisure activities accompanied by landscape
improvement in a phased development. After local government
reorganization in 1974, the new City Council published an Interim
Planning Statement for the valley which included a "Forest Park"
and a "Riverside Park,"  reflecting a more social  attitude to plan-
ning. These ideas eventually grew into the "Five Park Scheme," in
which the Enterprise Zone, designated by the government in 1980,
became the Enterprise Park  for industrial activity, accompanied by
a Leisure Park, a Riverside Park and,  in the old dockland area, a
City Park  and  a Maritime  Park (Fig. 5). Thus the concept of
restoration of the Lower  Swansea Valley has taken place within a
wider framework of urban renaissance than originally anticipated
by the pioneers  of the Lower Swansea Valley Project.
  Although small areas had been restored, no significant reclama-
tion took place until government assistance became available in
1966. Since that  date, reclamation has been active every  year,
culminating in  the last major scheme which began in 1983. The
reorganization of local government in 1974 gave  fresh impetus to
the process  of restoration as the new  Swansea City Council gave
priority to the work. A small executive committee was set up with
power to take decisions without going  through the normal lengthy
channels. As a result, full advantage has been taken of any finan-
cial aid, with the City Council able to respond rapidly and posi-
tively to any opportunities for redevelopment.
  The rate  of reclamation was controlled by the speed of land ac-
quisition by the City Council as the Welsh Development  Agency
would only accept grant applications for land which was in public
ownership. All the land recommended by the 1967 Report has now
been acquired by the City Council. One advantage of the publica-
tion of the earlier report was that land owners accepted the value of
unified ownership for redevelopment purposes. Reclamation began
with sites on the periphery of the valley, where there was access for
heavy machinery, and gradually worked toward  the center of the
                                                                                    INTERNATIONAL ACTIVITIES
                                                          555

-------
                                 EntgrpriM Park

                                 lenuce Park

                                 Riv*rsKh Park

                                 City Park

                                 Maritime Park
                           Figure 5
            Swansea City Council's Five Parks Scheme
                        Figure 6
Reclamation Schemes in the Lower Swansea Valley, 1966-1982
                                                              Table 1
                                       Reclamation Schemes in the Lower Swansea Valley, 1966-1982
                                                     (Bromley and Morgan, 1983)

MAP
COPE
a
A


B





C

D

E


F.


F


NAMT. OF SCHKMK


White Rock


Upper Forest and
Worcester Works
Phase I (Clearance)
Phase II (Filling.
Borrow from White
Rock)
Cwm , Winsh Won jn>i
Llansantlct
Dyffryn Works

Swansea Canal
Phase I (Drainage)

:~w.irv>^,i Cannl
Phase II (Filling)

Morfa I (Borrow
for .Swansea C.inal

1XVT

b
S


FZ





r.7.

F.V

N
an'J
S
N

AI;K»

ha
33


16





104

8

1 1


4

and I
*
[
)
u/.'i'h •:. UK wo:<>
start Finish



1907
Apr i 1
1967


1968

.i.m
1l/o'*
rv t.
1969
Aug
1970

NVv
1 c. 70



Nnv

1968
Dec
1967

Dec
1968

J.in
1970
July
1970
rvt-
iy7i

Due
1971



COST

c
Ell 3.917


ear, .391


C21.141 D


f 10,447
E
t 1 10.749
t
Ł10.172


C1 31 ,6rM





DUDGKT
YKAR

67/68

68/69
67/68


68/69


69/70
70/71
69/70
70/71
70/71
71/7^

70/71
71/7?




LAb

d
C


C


C


c

c

c


c





GRANT

e
8'.\ front Welsh Office


85% from Welsh Office


M » •


85\ from Welsh Office

CS9.272 from Welsh Office

85\ from Welsh Office


E90.189 from Welsh Office
and E?0,000 from British
Waterways Board


556       INTERNATIONAL ACTIVITIES

-------
Table 1 (continued)
MAP
CODE
a
G



H

I

J

K

L




M



N



0









P




Q





NAME OF SCHEME


Plasmarl: Cohen
Land and Graig
Brickworks
(Outside LSV)
Rose and Spelter
Works
Morfa II

RTZ I

Glamorgan Works

RTZ II
( includi ng
construction of
culvert costing
c. Ł53,000)
RTZ III and IV
Phase I


Morriston Lower
Gas Works


Upper Bank
Phase I (Re-
clamation and
laying sewer prior
to construction of
Athletics track)
Winsh Wen Earth-
Worka (Site
preparations not
derelict land)
RTZ III and IV
Phase II (Re-
clamation of Site 6
and culvert on
Site 4 )
Glandwr /Morfa
(Borrow for 9ite 6
reclamation)
NB Frederick Place
also borrowsite for
Site 6
LOC

b
N



EZ

S

EZ

EZ

EZ




EZ



EZ



S





EZ



EZ




S





AREA

ha
3



16

2

3.5

0.5

5.5




24.5



13



8.5









8.3




10.4





DATES OF WORK
Start Finish

Feb
1972


Feb
1974
June
1974
Oct
1974
April
1975
Nov
1976



April
1978


Oct
1978


Feb
1980




May
1980


Oct
1900



Oct
1980




Feb
1974


Mar
1975
Jan
1975
June
1975
June
1975
Sept
1977



Dec
1979


Mar
1980


May
1980




Oct
1981
and
after
April1
1982



April
1932




COST

c
Ł21,637
t


E105,455T

Ł110,204

Ł58.084

Ł4,845

Ł193,321
E
Ł122,555
Ł70,766

Ł271 ,619E

Ł268,300
Ł3.319
Ł173,266

Ł112,200
Ł61 ,066
Ł70,000





E135,000E



Ł502.000E

Ł283.000
Ł219,000







BUDGET
YEAR

71/72
72/73
73/74

73/74
74/75
74/75

74/75
75/76
75/76



76/77
77/78



78/79
79/80


78/79
79/80
79/80
80/81




80/81
81/82




80/81
81/82







LAB

d
C



C

C

C

C

C




C



C



C





D



C










GRANT

e
Ł4,717 from Welsh Office



—

Ł13,044 from Welsh Office
(LEA)
-

-

100% from WDA




100% from WDA



100% from WDA



100% from WDA





Ł40,500 from ERDF



100% from WDA










                           INTERNATIONAL ACTIVITIES
                                                            557

-------
                                                        Table 1 (continued)
MAP
CODE
a
R
S
NAME OF SCHEME
Upper Bank
Phase II
(Some borrow for
Site 6
Site 8c and
Site 14
I.OC
b
S
F.Z
AREA
ha
15

DATES OF WORK
Start Finish
Nov
1981
Jan
1982
April
1982
March
1982
COST
c
Ł237.811
(up to
Aug 1982)
Ł7.031 F
BUDGET
YEAR
81/82
82/83
81/82
LAB
d

C
GRANT
c
100% from WDA
100% from WDA
  NOTES:
  a. See Figure 9
  b. EZ •  Enterprize Zone
    N - Northern LSV outside EZ
    S - Southern LSV
  c. The total cut for each scheme is the first figure ipedfied. Where possible the costs are
    allocated to particular budget yean using the SCB and SCC Annual Budgets. Eicepl for
    scheme R. the costs are for the period prior to 31 March 1982. The tources of information are:
    C. PGPC 7 December 1967
    D. PGPC 6 January 1969
    E. SCC, Engineer's Department
    P. SCC, Planning Department
    T. SCC, Treasurer's Department

derelict  area. The sequence of reclamation schemes  is given  in
Table 1  and Figure 6, the work being done under contract by out-
side contractors but administered and supervised by the City Coun-
cil Engineering Department.  Schemes have  been initiated only
when financial assistance has been assured and the progress reflects
the provision of government funds.'
   It is difficult to  specify the exact cost of a reclamation scheme.
The costs listed in Table 1 are those for reclamation alone. They do
not include the preliminary cost of the land purchase, the costs of
administration and design or the subsequent landscaping expenses.
   The cost of reclamation schemes in the Lower Swansea  Valley
reflects the quantity and character of the material that  needs to be
removed or remolded. These  factors and the variety  of physical
problems presented by the industrial wastes and derelict buildings
have been summarized by the City Engineer.' The toxic character
of much of the tip material has been a particular problem." The
discovery of unforeseen obstacles has often resulted in increased
costs, such as the exposure of a brick-work culvert during reclama-
tion of the Rio Tinto Zinc  site.  Changes in policy regarding the
subsequent use of the reclaimed  land, and hence changes  in the
gradients of slopes required, have also led to increased costs. The
initial tender of Ł106,736 for clearance at the Rio Tinto Zinc site
was considerably below the final cost of the scheme, because an ad-
ditional  Ł30,000 was required for regrading alone. Frequently, it is
not until clearance is actually underway that the final decisions are
made on the required gradients at a particular site.
   The costs  of reclamation schemes have been reduced  by the
presence of recoverable materials on the clearance  sites, and by
undertaking schemes as combined projects. The Glamorgan works'
site, for  example, included slates and stone of value to the contrac-
tor,  and the existence of those materials reduced the  cost  of the
contract. Other schemes have produced usable hard-core. Schemes
conducted as combined projects have involved one site providing
fill for another. Thus, material from the Rose Spelter works' site
helped to raise the  level of an adjacent site (Items G and H, Table
1). When the reclamation is completed, the sites often have con-
siderable value for industrial development; this not only has  re-
duced the need and eligibility for a grant in some cases (as at Rose
Spelter works)  but  also has yielded occasional revenue from land
sales.
   Schemes  have   usually  been  initiated  only when  financial
assistance has been ensured so that the speed of reclamation has
reflected the availability of government funds. The schemes of the
late   1960s, when  Welsh  Office  grants  first  became available,
resulted  in the reclamation of 161 ha of land. The reclamation of
the White Rock tip involved work outside the valley floor, but is in-
cluded in Table  1  because material from it was used to raise the
                                                           d. C • Contract Labor
                                                             D - Direct Labor
                                                           e. The amount of grant specified is the amount awarded up to 31 March 1982
                                                           SOURCES:
                                                            SCC. Efigineer'i. Planning and Treasurer's Department: PGPC. PC and PRC minutes
                                                           lŁ- SI.JO (US)
                                                           level of the Upper Forest and Worcester works' site before con-
                                                           struction of the Morganite  factory. In the first  four years of the
                                                           1970s, only 18 ha were reclaimed; progress quickened in the second
                                                           half of the decade.
                                                                                       Industrial/commercial use

                                                                                             Derelict land

                                                                                          Vacant, reclaimed land

                                                                                             Opentpace

                                                                                             Roughland

                                                                                             Marsn

                                                                                             Farmland

                                                                                             Forest /Woodland
                                                                                      Figure 7
                                                                      Land Use in the Lower Swansea Valley, 1982
558
INTERNATIONAL ACTIVITIES

-------
  The interest of prospective developers has had a less significant
impact on the sequence of reclamation schemes, and most schemes
have been  undertaken  without a particular  developer in view.
However, the clearance and subsequent raising of the Upper Forest
and Worcester works' site was effected with the Morganite factory
as the expected after-use. Site preparation at the Winsh Wen site
was effected during late 1980 and early 1981 in order to make
several hectares of industrial land ready for the Enterprise Zone.
  The toxic wastes have been dealt with by encapsulation to avoid
the removal of large quantities of material out of and into the
valley, and the more benign steel waste material has been used to
transform the central part of the valley into a fresh landform upon
which factory units can be established. The present pattern of land
use is shown in Figure 7.
  From inception to completion, the restoration of the valley has
taken 25 years. Most of the scars of former dereliction have now
been removed, but shortage of suitable covering materials means
some areas  of copper slag are still to be covered; excavations for an
amenity lake are also in progress. Physical restoration may have
taken place but,  despite government  assistance,  the  economic
resuscitation of the valley is accomplished slowly at a time of na-
tional and worldwide recession.

ACKNOWLEDGEMENTS
  The author is grateful to Drs. Rosemary Bromley and Richard
Morgan  for permission to use material from their publication,
Changes and Industrial Development in the Lower Swansea Valley.
Other material in  this  contribution has been derived  from the
author's Surveying Derelict  Land,  to  be  published  in 1985  by
Clarendon  Press, Oxford. G.B. Lewis, Andrew Lloyd  and Gary
Llewellyn drew the illustrations.
REFERENCES

 1.  Hilton, K.J., ed.,  The Lower  Swansea Valley Project, Longman,
    London, 1967.
 2.  Street, H.E. and Goodman, G.T., "Revegetation techniques in the
    Lower Swansea Valley."  The Lower Swansea  Valley Report, K.J.
    Hilton, ed., Longman, London,  1967.
 3.  Weston,  R.L.,  Gadgil, P.D.,  Salter, B.R. and  Goodman, G.T.,
    "Problems of revegetation in the Lower Swansea Valley—an area
    extensive industrial dereliction." Ecology and the Industrial Society,
    G.I.  Goodman, R.W. Edwards and J.M. Lambert, eds., Blackwell,
    Oxford, 1965.
 4.  Bridges,  E.M., "Eroded soils  of  the Lower  Swansea Valley,"
    Journal of Soil Science, 20, 1969, 236-45.
 5.  Bromley, R.D.F. and Humphrys, G., eds., Dealing with Dereliction,
    University College of Swansea, 1979.
 6.  Bridges, E.M., Chase, D.S.  and Wainwright, S.J., "Soil and plant
    investigations since 1967." Dealing with Dereliction, R.D.F. Bromley
  and G. Humphrys, eds., University College of Swansea, 1979.
 7.  Bridges, E.M., Chase, D.S. and Wainwright, S.J., "Distribution of
    copper, lead, zinc, cadmium and nickel in plants, superficial layers
    and mineral soils of the Lower Swansea Valley." The Productivity of
    Restored Land, Land Decade Council, London, 1981.
 8.  Bromley,  R.D.F.  and Morgan,  R.H.,  Change and Industrial  Re-
    development in the Lower Swansea Valley,  University College of
    Swansea, 1983.
 9.  Jones, H.I.I.,  "Engineering problems and their solution." Dealing
    with  Dereliction,  R.D.F.  Bromley and  G. Humphrys, eds., Uni-
    versity College of Swansea, 1974.
10.  Chase, D.S.  and Wainwright, S.J., "Vertical distribution of copper,
    zinc,  lead ions in weathered tips of copper smelter waste in the Lower
    Swansea Valley," Environmental Pollution (Series B), 1983, 133-46.
                                                                                       INTERNATIONAL ACTIVITIES
                                                           559

-------
       INVESTIGATION  OF  LAND AT THAMESMEAD  AND
       ASSESSMENT  OF REMEDIAL MEASURES TO BRING
           CONTAMINATED SITES  INTO  BENEFICIAL USE
                                            GEORGE W.  LOWE
                                          Scientific Services Branch
                                          Greater London Council
                                  County Hall, London, United Kingdom
INTRODUCTION

  In the mid-1960s, more than 1,000 acres of land (a major part of
the Royal Arsenal in Woolwich) were obtained from the Ministry
of Defence by the Greater London Council for redevelopment. At
that time, it was the biggest single area of vacant land in London.
With other land at nearby Abbey Wood, the acquisition provided
an opportunity for creation of  the new community of Thames-
mead.
  For more than 200 years, the western end of the site was sub-
ject to intensive use for the manufacture of armaments. With the
growth of technology,  an expansion of industrial activity  took
place, reaching its peak in the  war  of 1914-18. Expansion  took
place eastwards along the Thames: as output increased, vast quan-
tities of waste  were generated by industrial processes, manufac-
ture of town gas, generation of electricity and the testing of ex-
plosives.
  Large mounds  were  deliberately created from waste to form
cover for sensitive installations, and marshland was filled with in-
dustrial residues to provide foundations for buildings and a net-
work of rail tracks and roads. Following the end of the  Second
World War, marked by the destruction by burning of great quan-
tities of surplus explosives and disposal of bomb damage  rubble,
the installations gradually fell out of use, leaving a legacy of dere-
liction that provided only little evidence of a new and potentially
hazardous environmental problem.
  Throughout the older districts  of Greater London, the industrial
contamination of land emerged  as a form  of pollution not prev-
iously seen as a major cause of  concern. However, with demand
for inner city development land  coinciding with the decay of tra-
ditional  industries, it followed  that housing and schools would
have to be built on land not previously considered for such use.
The condition of ex-industrial land has demanded careful examina-
tion before redevelopment, and the Royal Arsenal lands are no ex-
ception.
  A multi-disciplinary team was  assembled to investigate the prob-
lem of contaminated land at Thamcsmead, and their work  con-
tinues. The team's work is part  of a comprehensive approach to
reclamation which includes the creation of new land by controlled
deposit of contaminated material and the profitable recycling of
surplus excavated materials from elsewhere.

THAMESMEAD SITE

  The development area of Thamesmead is 16 km from the center
of London. It covers approximately 650 ha and is situated on the
South Bank of the Thames with a 5.5 km frontage to the river. The
land is composed  almost entirely of drained marsh at the  foot of
rising parkland which forms a backdrop to the initial stages of new
development.
  Much of the land developed to date has no known history of in-
dustrial usage and has not been subject to contamination. The re-
mainder, which formed  a major  part of the Royal Arsenal, is 400
ha at the North of a raised embankment containing the London
Southern Outfall.
                                                   Ground Conditions and Drainage
                                                     The ground is composed of silty clay up to 1.5 m thick overlying
                                                   peat and alluvium 5 to 7 m deep. At the base of the alluvium is a
                                                   stratum of water bearing gravel on Thanet sand and chalk.
                                                     The site forms a natural drainage basin between the Thomas and
                                                   the Wickham Valley which lies at the base of the rising ground to
                                                   the South. Historically, surface water run-off from the hills drained
                                                   into the area by way of culverts; after flowing naturally across the
                                                   site through a network of ditches, the water was discharged into the
                                                   Thames via tidal sluices. Typical of marshland, the site is flat and
                                                   naturally featureless except for trees.
                                                     The high water  table and poor  load bearing capacity of the
                                                   ground require that all construction work be piled or surcharged.
                                                   The Tilling of undrained marsh in earlier times has led to many of
                                                   the present pollution problems due to the industrial origin of the
                                                   materials used for reclamation.
                                                   INDUSTRIAL HISTORY

                                                     The original river bank was probably built up by the Romans,
                                                   and there is evidence of a settlement at nearby Plumstead. In the
                                                   12th century, an Augustinian Abbey was established on  the rising
                                                   land to the South. From that time until 1524, the monks from
                                                   Lesnes played a part in reclaiming and draining the land.
                                                     Following a period of neglect during which severe flooding was
                                                   experienced, the responsibility for maintaining the river wall and
                                                   draining the marshes passed to Commissioners for Sewers in about
                                                   1600. During the reign of Henry VIII, industrial development of
                                                   the area was foreshadowed by the establishment of the Woolwich
                                                   Naval  Dockyard. Downstream,  land destined to become part of
                                                   Thamesmead known as "The Warren" became a naval and  mili-
                                                   tary center in the 17th  century when Ranges were established and
                                                   storehouses for guns were built.
                                                     The earliest major manufacturing industry in  this area  started
                                                   with construction  of a brass foundry by Sir John  Vanbrugh in
                                                   1716 when the first cannon were cast. The output of the foundry
                                                   grew, and the Royal Regiment of Artillery moved its barracks to
                                                   Woolwich in  1719 to be near its main source of supply. The Royal
                                                   Artillery depot remains in Woolwich to this day, as does the  orig-
                                                   inal brass foundry  which has been preserved as a building of his-
                                                   toric interest.
                                                     Land was reclaimed from the marshes as the establishment grew,
                                                   and land  filling became an occupation of convicts held in prison
                                                   hulks  anchored in  the river adjacent to The Warren.  Mud and
                                                   debris from  the construction of St. Katherines Dock near the
                                                   Tower of London  was taken to the site for filling purposes, and
                                                   this material was heavily augmented by waste products from the
                                                   gun foundry and other industrial installations. By the 19th century,
                                                   The Warren had been renamed "The Royal Arsenal." and a period
                                                   of major expansion began. By  1890, the Arsenal covered 324 ha
                                                   which included practice and experimental ranges. A further 160 ha
                                                   of marsh were added at this time and 11 moated magazines were
                                                   constructed,  some of  which are now being preserved for their
                                                   amenity value.
560
INTERNATIONAL ACTIVITIES

-------
              or neavy weapons, ammunition and explosive de-
vices reached its height in the war of 1914-18 when some 80,000
people were employed in armaments work. Coal and raw materials
entered the area via a spur from the adjacent railway, by barges
into the Arsenal dock or by ships using piers to deep water in the
river.
  The Arsenal was self supporting as it produced its own town gas
(from about 1850); later, a steam plant and electric generating sta-
tion were added to the facilities. The prodigious quantities of ash
and waste produced by these plants alone, over their long life, have
remained to make their contribution  to the topography of the
land and the problem of contamination.
  The arsenal continued with production and testing of modern
armaments through the war of 1939-45 when it proved to be vulner-
able to air attack.  A large number of bombs  fell on the area, and
over 1,000 people lost their lives. After the war, vast quantities of
surplus explosives were destroyed on the site by burning;  the prod-
ucts of the burning added to the growing quantities of waste ma-
terial  from  industrial/experimental  processes  and  debris  from
destruction of obsolete installations.
  The area finally was sold  to the GLC; it  was handed over in
stages during the 1960s. Withe the demolition of remaining build-
ings, the stage was set for the development of Thamesmead, which
at that time was proposed to be a town for 60,000 people.

THE DISCOVERY OF INDUSTRIAL CONTAMINATION

  Development of Thamesmead started in an area to the south of
the Arsenal boundary, and it was some time before work of any
magnitude began on land where there had been any concentrated
industrial activity.  Not until 1975 was the extent of industrial con-
tamination fully revealed, although its presence had been suspected
for some time. The initial impact  on site development was  spec-
tacular, because the discovery  of  heavily contaminated material
during  excavation  work was  accompanied by a chemical fire and
the penetration of buried tanks containing liquid residues from the
former gas works.
  The situation was considered sufficiently serious to suspend con-
struction operations to avoid risk to personnel. The financial pen-
alty for stopping  the contract prompted  a project-wide inquiry
which indicated that all land used by the Arsenal must be thorough-
ly examined. Hazards to health had  to be assessed and means of
providing  protection  had to be determined,  both for the  work-
force and for eventual users of the land.
  Significantly, the problems at Thamesmead began to emerge at a
time when similar difficulties  were being encountered by the Coun-
cil on other sites in London and by other development authorities
elsewhere.

BACKGROUND TO THE INVESTIGATION
  The working party of officers established to investigate and deal
with the problem  were faced with a number of difficulties, not
least of which was the  obviously diverse but classified nature of
former industrial activity.  Buildings housing industrial  processes
had  been  constructed  on marshland  reclaimed with industrial
waste and other fill materials of unknown origin. Furthermore,
the practice of using industrial waste for foundations to the Arsenal
network of railway lines, vehicle tracks and as protective cover to
sensitive installations led to a land make-up of some complexity.
  Use of rubble from bombed buildings in East London as backing
to the river wall was not necessarily a cause for concern, but waste
disposal from the Arsenal at the same time and in the same loca-
tions led to random localized filling patterns.
  Destruction by burning  of war surplus explosives, incendiary
devices and other  dangerous material on areas of unused marsh
and industrial waste sites again tended to complicate the investi-
gation.
  In contrast to the mixed and apparently serious contamination
which was sometimes evidenced by severe distress to vegetation,
the Eastern sector of the site was typical of pastureland on drained
marsh supporting a healthy plant  growth and  habitat for wild
life.
   Conversely, it became clear that isolated areas of severe contam-
ination existed in the generally clean, Eastern part of the Arsenal,
while in the heavily industrialized West and Central parts there
remained zones of thickly wooded clean land.

SCOPE OF THE INVESTIGATION

   The complexity of land use and random disposal of waste neces-
sitated a total examination of the site previously occupied by the
Royal Arsenal. Apparently clean areas would also be subject to in-
vestigation as there was evidence to suggest that a seemingly un-
disturbed ground surface was not a  reliable indication of contam-
ination-free conditions beneath.
   The determination of site specific sampling strategies depended
very largely on the following:
•Such history of the site to be tested as is possible to collect
•A visual inspection of the ground
•Details of the intended land use
   The last item intended  use of the land, ultimately determines
the scale of long term risk; houses with gardens being at the top of
the scale. The first two items provided  an indication of the con-
dition and previous usage of the ground. Based on this initial eval-
uation, a sampling  strategy was devised with sampling  intervals
which have used trial pits on 100 m centers (rarely) and as close
as 10 m  (center  to center) where conditions  are considered to be
particularly bad. Usually, trial pits were excavated on a 50.0 x 25.0
m staggered grid.
   As stated previously, the original  use of long demolished build-
ings is difficult to determine, but it has been possible to identify
the following activities.
•Heavy machine shop and forging work
•Non-ferrous metal foundries
•Cadmium and other metal plating
•Town gas manufacture
•Development and testing of paints
•Manufacture of acetylene
•Manufacture and testing of weapons and explosives
•Destruction of surplus explosives by burning
•Destruction of surplus incendiary devices
•Storage of coal stocks on the surface
•Dumping of industrial waste
   Such activities have prompted a fairly broad spectrum of soil
testing and, as investigation has progressed  to suit  the  program
of development,  the list of analytical determinations has been mod-
ified by experience gained. From the outset, soil analysis has been
carried out by external consultants to meet a fluctuating and often
heavy laboratory workload.  Determinations needed  and  methods
of testing are specified by the Scientific Adviser to satisfy the needs
of the Medical Adviser in his assessment of health  hazards, and
to provide data on possible deterioration of substructures.
  Analytical determinations  most commonly called  for are indi-
cated  below, but variations are made to suit particular circum-
stances:
•pH
•Acid Soluble Sulphate
•Magnesium
•Elemental Sulphur
•Sulphide
•Total Cyanide
•Free and Complexed Cyanides
•Phenols
•Toluene and Cyclohexane Extractable  Material
•Coal Tar and Mineral Oil
•Total and/or Available Metals: Lead, Cadmium, Mercury, An-
 timony, Arsenic, Zinc, Nickel,  Lead,  Copper

SOIL SAMPLING PROCEDURES

   When the sampling strategy is agreed, the coordinator arranges
the following:
                                                                                    INTERNATIONAL ACTIVITIES
                                                          561

-------
•Preparation of a sampling drawing based on a 1:500 scale survey.
 The drawing indicates the operational grid for Thamesmead, and
 trial pits are related to this for ease of setting out. Each pit is given
 a code name reference to enable easy identification of location
 on-site.
•Surveyors set out the trial pits, locating each with a code-marked
 driven stake
•The engineer responsible for site investigation employs a contrac-
 tor for excavation  of trial holes by JCB backhoe excavator and
 provides an officer for his supervision. A scientist directs the oper-
 ation and ensures that samples are taken to a set routine.
•Soil samples are taken at established vertical intervals in the trial
 holes usually down to 3.0 m. If deeper holes are required, a track
 driven excavator is employed.  Bore holes are  not favored unless
 very deep samples are required.
•Samples are placed in scalable 1 1 plastic buckets identified by
 previously coded self-adhesive labels.
•Pre-coded log sheets are filled in as sampling proceeds, being care-
 ful to use a standard nomenclature  giving  the most  accurate
 possible description  of the  sample  including transient  features
 such as color and smell.
•A log sheet of each trial  pit is prepared describing the form and
 thickness of ground strata. When compared with other trial pit
 data, this is used to identify horizons of polluting material.
•Samples are batched and sent to a consultant analyst  together
 with log sheets and  any special instructions which the scientist
 deems appropriate.

WATER TESTING
  The water regime at Thamesmead is complex and dynamic. The
general water table is high and fluctuates due to the site being in a
drainage basin. The adjacent Thomas is tidal, and the river level is
often considerably higher than that of the land. There is also a
"perched"  water table  some  distance below the surface.  As
Thamesmead grows, the surface water drainage system is extended
with the canal network, and the holding capacity of the balancing
lakes is increased.
  As  indicated  earlier, Thamesmead has been  reclaimed  from
marshland, and  successive bank raising and land drainage opera-
tions have led to the present surface water and flood control sys-
tem which  depends on canals,  balancing lakes and tidal sluices,
plus a major pumping station for emergencies. This system is being
constantly extended, and undertaking general water testing at the
present stage would be inconclusive, bearing in mind that a great
deal of major earth  moving is still to be done in both heavily con-
taminated and relatively clean areas. As a consequence, only  rou-
tine monitoring is undertaken in situations where the public health
needs of the local population have to be safeguarded. In the longer
term, it is intended to complete a comprehensive survey of the com-
plete water regime.

METHANEGAS
  Methane gas generation below the surface  has  been  a major
factor in a number of development schemes on sites bordering the
Thames where  land has been reclaimed from the foreshore. In
such cases, the generation of gas is usually associated with break-
down of organic material  in river silt which has been enclosed by
river wall construction  and covered  by landfill material.  Where
bodies of silt are large, generation of methane gas has been sub-
stantial,  demanding expensive  measures in the construction of
buildings to  avoid  the  possibility  of gas  accumulation and ex-
plosion.
  The dominating feature of Thamesmead is  the River  Thames,
and it  has been necessary to consider the possibility of  methane
generation, particularly on sites close to the river wall. Fortunate-
ly,  there is only one area  of land where the level of methane gas
has been found significant, and this has undoubtedly arisen from
the enclosure of a relatively small "bay" of foreshore silt during
localized bank raising in the last century.
  Living accommodations are to be  built  on the  site, and it has
                                                        been possible to make recommendations at an early stage in the de-
                                                        sign process which will enable the design of visually and function-
                                                        ally acceptably built structures without compromising the overrid-
                                                        ing need to avoid gas accumulation. Recommendations include the
                                                        introduction of lateral dispersal arrangements using granular break
                                                        layers in combination with suspended ground floors without foun-
                                                        dation beams and,  at the other extreme, the placing of all living
                                                        spaces on upper floors with parking or garages on the ground.

                                                        FIBROUS ASBESTOS DEPOSITS
                                                          A search for deposits of fibrous asbestos is being carried out
                                                        simultaneously with the investigation for industrial contamination.
                                                        This investigation has been necessitated by the discovery of asbes-
                                                        tos insulation on the surface, along the lines of a long removed net-
                                                        work of overhead steam and hot water pipes which extended for
                                                        several kilometers across the site from each of two boiler plants.
                                                        This form of heating was necessary to avoid the risk of fire. The in-
                                                        sulation  contains both crocidolite and  chrysotile and must be re-
                                                        moved before site clearance and construction work can take place.
                                                        A 10 m  x 10m search  pattern is being carried out based on the
                                                        100 m operational grid. It is a tedious process  for the personnel,
                                                        but there is no other practical way of locating asbestos deposits in
                                                        densely overgrown conditions.
                                                          Clearance  of each area is undertaken by licensed  contractors
                                                        who originally removed the asbestos off-site to official disposal
                                                        dumps. Recently, more certain methods of disposal have been put
                                                        into operation at Thamesmead using on-site licensed disposal facil-
                                                        ities.
                                                        SPONTANEOUS COMBUSTION
                                                          Foundry work, the production of town gas, generation of elec-
                                                        tricity and the generation of steam for industrial processes and dis-
                                                        trict heating  demanded  the import of great  quantities of coal. A
                                                        deep water pier capable of dealing  with  ocean going colliers is
                                                        located at the western end of the Arsenal area, and a spur from the
                                                        nearby railway was used to bring in the coal from mines in Kent.
                                                          Presumably, to ensure continuity  of provision during the last
                                                        war, stocks of coal were held on the surface in locations behind the
                                                        old river embankment.  Some of the coal sank  below the surface
                                                        and remained after the main stockpiles were removed for normal
                                                        use.
                                                          During an unusually hot and dry summer in  1976, two sites of
                                                        spontaneous  combustion became manifest when a belt of  trees
                                                        which had colonized the area died and fell to the ground as  their
                                                        root systems were burned away. In this case, the fires were allowed
                                                        to burn through the remaining coal, but all trees and flammable
                                                        materials were removed from the surface to reduce the fire loading.
                                                        The fire  burned out over a period of 2 yr leaving a residue of ash
                                                        and debris which has since  been recolonized  by scrub bushes.
                                                        Prior to  the  development which is to commence shortly, the site
                                                        will be excavated down  to natural clay and then filled with clean,
                                                        imported granular material.
                                                          In another situation,  spontaneous combustion on a raised em-
                                                        bankment was impossible to control by conventional means. As the
                                                        adjacent land was required  for early development, the only eco-
                                                        nomic solution was to remove the entire embankment with the fire
                                                        brigade  standing by to prevent spread of fire  and  to  wet the
                                                        material  as it was excavated.
                                                          Both incidents proved to be valuable. Much useful data  were
                                                        obtained during site tests and  laboratory analysis of the recovered
                                                        combustible  material. For guidance purposes,  it has been deter-
                                                        mined that sites containing material  below the surface with  a
                                                        calorific  value of 1700 cal/g have a potential  for spontaneous  com-
                                                        bustion.

                                                        SITE CONDITIONS AND REMEDIAL MEASURES
                                                          The working party of officers established to deal with land pollu-
                                                        tion at Thamesmead eventually became the  Council's Assessment
                                                        Panel on  Contaminated  Land  which  now  has a London-wide
                                                        responsibility with respect to all OLC development. The Panel is
562
INTERNATIONAL ACTIVITIES

-------
interdepartmental and  multi-disciplinary and consists mainly of
officers not engaged on contaminated land full time.
  The areas of professional activity covered by the members of the
Panel are as follows:
tEnvironmental science and analytical chemistry
•Structural engineering and statutory building control
•Environmental health procedures, health and safety
•Horticulture
•Hazardous waste disposal
•Public Health Engineering
•Construction and management (coordination)
  The objectives of the Panel in making their final assessment of
investigation results are briefly as follows:
•To determine the condition of the site
•To identify and express the main areas of risk
•To consider and recommend remedial  measures  with alternative
  solutions if possible
•To outline minimum requirements for the protection of construc-
  tion and other workers
•To determine the category of excavated material for disposal
•To initiate further inquiries if necessary on deterioration  of sub-
  structures and to consider anti-explosion measures for buildings
  when methane is a problem
•To recommend  suitable forms  of landscape  treatment  and
  methods of tree preservation
•To arrange long term storage and retrieval of data from the inves-
  tigation
   The staff rely heavily on a detailed report prepared by  the en-
vironmental scientist which includes and is based upon soil analy-
sis, trial pit and strata logs,  drawings indicating the site sampling
patterns and supplementary reports on combustibility and methane
generation where appropriate.
   The present condition of the land is evaluated in the  context of
 its intended development so that remedial recommendations appro-
 priate to the proposed use can be proposed. Long term hazards to
 the health of land users is the primary consideration,  and this sets
 the pattern for remedial work. Occupants of houses with gardens
 where fruits and vegetables may be grown  are regarded to be most
 at risk; as the bulk of development is taken up by housing, most
 attention is directed to this concern.
   Schools are given particular consideration, especially where there
 are  possibilities of horticultural activity, but generally the build-
 ings and hard play surfaces are seen to be satisfactory forms of
 protection in themselves. The bulk of remedial measures are, there-
 fore, included in landscaped areas.
   As indicated earlier, some sites which are grossly polluted  sup-
 port mature  and  apparently healthy trees, and  where possible,
 these trees have been saved by a special  localized technique which
 avoids overfilling to the point where they might die. The technique
 demands careful work using hand tools to remove soil from be-
 tween roots and then replacement with  selected  material. A  gen-
 eral covering of no greater depth than 0.2 m is applied to  prevent
 demage and decay at the base of the tree trunk. Under no  circum-
 stance, however, would this step be undertaken in high risk areas
 which could not tolerate a reduction in capping depth.
   The many forms of development at Thamesmead, ranging from
 housing to industry, demand that each site be dealt with on its
 merits, and experience has proved that the condition of the land is
 as varied as its many uses.
 SHORT TERM AND INCIDENTAL RISKS
   Through the  Medical  Advisor's Chief Environmental  Health
 Officer, the Panel determines if there are any special requirements
 with respect to the health and safety of site construction personnel.
Fortunately, with few exceptions, the condition of sites at Thames-
 mead has  not prompted special needs beyond the use or ori-nasal
masks and/or water  spraying at most. The mandatory provisions
of washing facilities and accommodation for consumption  of food
under the Health and Safety at Work Act for site personnel are
satisfactory for the majority of sites examined so far.
  Incidental risks will occur when repairs or alterations are made in
the longer  term to buried services. Low voltage electricity, tele-
phone and  television cables are usually buried about 0.5  m below
the surface; HV cables, water and gas mains are laid with not less
than 0.9 m cover; and drainage is usually deeper still.
  It would  be unrealistic to assume that  public  utility or local
authorities  will be kept aware of possible health risks in the longer
term, so where necessary, recommendations are made for  trench
backfilling  to  take a form which reduces the  risk  of contact and
spread of foul material over clean surfaces  during reexcavation
for repairs, etc. Consequently, the  surplus  excavated  material
from all trenches intended for buried services is removed from site
and clean material is provided for backfilling wherever the capping
layer on a reclaimed site is likely to be penetrated.

RECLAMATION

  The methods of reclamation recommended by the Assessment
Panel have been implemented  to  resolve  contamination by the
simplest possible means:
•Overfilling with clean imported material
•Excavation and removal of contaminated soil and replacement
 with clean imported fill where  the original ground level has to be
 maintained
•The covering of sloping surfaces when the contaminated inner
 cores of raised mounds are exposed or deep cuts are formed for
 canal construction through raised areas
•Special arrangements for prevention of direct contact and erosion
 when excavations are lakes bring  subsurface  contamination into
 conflict with open water
•Change of land use
  Filling of large areas of land  poses many difficulties in locating
supplies of suitable material and transporting it to the site. If, for
instance, a site of 16 ha is filled to a depth of 1 m, the import of
160,000 m of material involves possibly 20,000 vehicle movements
between site and source of fill. This traffic is clearly undesirable in
an urban area.
  Ideally, the material to best deal with heavy metals and, indeed,
many other contaminants, would be clay; this is not always satis-
factory as a working base for construction and could lead to prob-
lems later due to its inherant instability in changing weather con-
ditions.
  Estuarine sand containing  approximately 10% silt or pit dug
sand with the same silt content could provide the desired protection
to deal with metals, but experience with such material has not been
satisfactory. Dredging  for estuarine  sand of the correct constit-
uency may not be permitted at  the time it is needed for develop-
ment, and in any event, the unloading of vessels by conventional
means which would be essential to preserve  the silt content,  would
be inordinately expensive.
  Put dug  sand providing a minimum silt content of 10%  is ob-
tainable, but a consistent upper limit of silt in local sources is vir-
tually impossible to achieve. Such material has been used already
at Thamesmead and found wanting  because excessive silt  in the
sand causes the site to become a  quagmire during bad weather.
Hydraulic Fill Using Sea Sand

  A large part of the Western area of Thamesmead has been filled
with hydraulically placed sea dredged sand for construction pur-
poses. The sand is  taken  from the North Sea by dredger and
pumped  ashore using river water  as  a transport medium  in the
proportion of approximately 10 to 1  by volume. Water is drained
back to the river leaving the sand which is sharp and contains a lot
of shell, but little or no silt.
  This method, previously regarded  as unsatisfactory for dealing
with contaminated land due to  the lack of  fines, is being adopted
for capping some areas by taking into account the additional thick-
ness of surface soil necessary to create a suitable environment for
plant growth.  Extra top/sub-soil is necessary for shrub planting
and large tree pits are essential, but  the  advantages of the fill
method are considerable, particularly in respect of its lack of en-
                                                                                      INTERNATIONAL ACTIVITIES
                                                           563

-------
vironmental impact on surrounding areas. Sites are self draining
and the material provides a clean working platform for construc-
tion work. Topsoil is placed late in the construction period.
  Other departments in the Greater London Council are dealing
with major development  schemes which are in no way associated
with Thamesmead.  Major road  schemes fall into this category,
and construction work often produces substantial quantities of sur-
plus excavated material which must be transported some distance
for disposal. By arrangement with the department concerned, it
has been possible to obtain good quality material  for reclamation
work at no cost when such projects are close to Thamesmead.
  Similar arrangements have  been made between general contrac-
tors in the South East London area and a schedule contractor em-
ployed by the Thamesmead  General  Manager. In all cases, the
source of material  is inspected and, where necessary, a full his-
torical check and sample analysis are made before acceptance.
More than 200,000 m' of material have been obtained  by these
means to date; these materials very often have been spread and
levelled on-site without charge.

The Arsenal Gas Works

  The grossly polluted condition of the old Arsenal gas works site
prompted the Assessment Panel to recommend a change of land
use whereby all building work within the area would be prohibited.
The ground has been saturated by liquid residues during three gen-
erations of gas  production  going back  to  approximately 1850.
These  liquid, accompanied by high levels of other contaminants,
have produced a below ground environment of considerable risk to
construction workers.
   It is fortunate that a nearby site intended for open space could
be relocated, leaving a more suitable area available to make up the
loss of building land. The gas works site has now been covered with
at least  1.0 m  of broken concrete and  rubble from demolition
works, followed by a layer of coarse shingle and then capped with
a 0.3 m layer of clay graded to falls and taken down at least 1.0 m
below ground at the perimeter. Open space and playing fields will
be established on the site, but trees will only be planted outside the
capping. The depth of subsoil/topsoil on the clay will be not less
than 0.5 m and will exceed 1.0 m in part.
  The opportunity  for this kind  of land exchange is rare and can
only be exploited in a situation  where area planning and project
management activities are undertaken comprehensively.

Corrosion Problems and Deterioration of Concrete

   As a matter of course, all sites are subject to  investigation to
establish their geophysical properties and natural levels of acid sul-
phates. On some sites at Thamesmead, excessive levels of acid sul-
phates have demanded special protection to piles, especially where
other concrete damaging materials are in close association. In such
cases, pile shells have been coated with 2-3 applications of an epoxy
resin-based coating.
   In most situations where pile  protection of this kind has been
necessary, it has arisen because of deposits of industrial waste used
to reclaim areas of marsh. Interestingly, very high levels of acid
sulphate (e.g., pH 3.5, Sulphase 5.4%) have been located on sites
entirely free  of industrial pollution. Similar results have been ob-
tained at depths of 2.0 m or more and  presumably arise from
natural decomposition of peat by microbial activity or (most likely)
the ingress of saline water from the Thames estuary in earlier times.

DISPOSAL OF CONTAMINATED MATERIAL

   The requirements of the Control of Pollution Act 1974 would
have placed  the development of a large part of Thamesmead in
jeopardy had it not been possible to provide on-site disposal facil-
ities for contaminated material from construction sites. Construc-
tion of a new river wall as part of the London flood defense scheme
created a space between it and the old river wall of approximately
350,000 m1. This space was licensed as a disposal facility suitable to
receive contaminated  material  excavated at Thamesmead. The
facility, which was  the first one of its kind in London, has proved
                                                        to be a major success in the rehabilitation of contaminated land
                                                        while providing an additional bonus in the reclamation of an area
                                                        of river bed without the need for importing material. The reclaimed
                                                        land, now suitably capped with clean material, will  be used for
                                                        housing development.
                                                          A  similar space has been created with the construction of an*
                                                        adjoining section of river wall, and a license has been granted for
                                                        the use of this for disposal of contaminated  material. In both
                                                        cases, the disposal license allows the burial of bagged fibrous asbes-
                                                        tos from  the cleanup operations described above.  Together, the
                                                        two sites  will provide more than 12 ha of riverside development
                                                        land.
                                                        RECLAMATION AND RECYCLING
                                                          Like many  sites  in the older  run down industrial areas  of
                                                        London and other cities, the land at Thamesmead is not only con-
                                                        taminated with the waste products of manufacture but also con-
                                                        tains the remains of generations of substantial buildings. Reclama-
                                                        tion and redevelopment normally demand the import  of clean fill
                                                        and export of excavated spoil and building rubble. It is unusual to
                                                        draw major development advantage from such operations. Because
                                                        of the size and form of the Thamesmead project, however, it has
                                                        been  possible to adopt a broad environmental policy  whereby all
                                                        excavated spoil is retained and concrete rubble convened for reuse.
                                                          Controlled use of contaminated material to create new develop-
                                                        ment land in conjunction with flood defense works is of consider-
                                                        able financial advantage to Council and developers alike. A similar
                                                        advantage has been derived from removal and crushing of old con-
                                                        crete structures to provide substantial quantities of foundation ma-
                                                        terial for roads and hardcore  for general construction purposes
                                                        virtually at the point of use. Combined with the ability to accept
                                                        unwanted  material from other developments for capping in con-
                                                        taminated areas, a comprehensive approach to reclamation and re-
                                                        cycling has proved technically and financially successful.
                                                        THEFUTURE
                                                          Much of the old Arsenal land still needs to be examined; part of
                                                        this is the estimated 1,000,000 m' elevated area of Tripcock Point.
                                                        This mound is covered in dense thicket and is composed almost en-
                                                        tirely of ash, industrial residues and gas works waste. It was shaped
                                                        to enclose a network of narrow roads and railways  between  steep
                                                        embankments into which explosives magazines and test facilities,
                                                        etc., all contained in massive masonry enclosures, were built. This
                                                        area  presents a special challenge because  as much of the present
                                                        physical form of the site as possible should be maintained to en-
                                                        hance its intended use as a District Park and a wildlife study area.
                                                          Altogether, some 80 ha  of Thamesmead still require investiga-
                                                        tion, most of which will demand an approach to reclamation which
                                                        takes into account unusual  physical difficulties. In one instance, an
                                                        elevated although flat site of approximately 23 ha is really a former
                                                        industrial  waste site where  waste was dumped on the remains of
                                                        more than 100 buildings demolished by explosives. It will not be
                                                        economical to recycle the concrete and masonry due to its depth be-
                                                        low the surface, thus  preventing the use of conventional driven
                                                        piles  for construction purposes. In this case, remedial  measures to
                                                        deal with pollution will have to be designed to meet additional re-
                                                        quirements for stabilizing the site.
                                                          In another instance, the  surface level of a site of approximately
                                                        35 ha will have to be elevated considerably for drainage purposes
                                                        using hydraulic filling methods. Care will  have to be taken to find
                                                        a suitable means of keeping water,  which  is pumped ashore as the
                                                        vehicle for fill displacement, out of contact with elevated berms of
                                                        contaminated waste which are immediately adjacent.
                                                          It is hoped the investigation  of the outstanding areas of land at
                                                        Thamesmead will be completed by the autumn of 1985, by which
                                                        time it will be possible to  reflect on the lessons learned over the
                                                        preceding 10 yr. So far, it  can be said that the Thamesmead site
                                                        is most noteworthy for the variety of forms of pollution and the
                                                        scale of the operation, rather than the intensity of contamination.
                                                        It is anticipated that this will continue.
 564
INTERNATIONAL ACTIVITIES

-------
  REMEDIAL ACTION FOR GROUNDWATER PROTECTION
         CASE STUDIES WITHIN THE  FEDERAL  REPUBLIC
                                           OF  GERMANY

                                                KLAUS STIEF
                                       Federal Environmental Agency
                                          Umweltbundesamt,  Berlin
 INTRODUCTION

  Remedial actions at abandoned landfills and contaminated in-
 dustrial sites are a growing challenge for water  and waste
 authorities and for scientists in the Federal Republic of Germany.
 Of  primary concern  are  harmful  impacts  on public  health,
 agriculture and groundwater.
  Past and current waste disposal practices and careless handling
 and storage of hazardous materials have resulted in contaminated
 soils, surface water and groundwater. While remedial actions are
 often designed to neutralize the hazards of contamination in situ,
 without touching or excavating the contaminated materials, field
 experience demonstrates that remedial actions are still more dif-
 ficult than safe handling and disposal of hazardous wastes in the
 first place.
 IN SITU TREATMENT OF
 ARSENIC CONTAMINATED GROUNDWATER   CtylOC^   A
  One of the earliest in situ German groundwater remediation pro-
 jects was carried out in Nievenheim, Land Nordrhein-Westfalen,
 near Cologne, between 1971 and  1979. From 1913 to 1971, about
 4,200 metric tons of calcium arsenate-containing sludges had been
 disposed of on the  site.  These sludges were  the by-product of a
 sulphurous  acid flue gas washing process in a nearby zinc ore
 smelter. Arsenic trioxide (As^) in the effluent had precipitated
 with calcium hydroxide (Ca(OH)2) at pH 8 as calcium orthoarsen-
 ate (CA3(AsO4)2).
  The groundwater had maximum concentrations of arsenic at 56
 mg/1, compared to a normal level of 0.01 mg/1. Sediment analysis
 revealed 10 to 170 mg As/kg, with an average of 78 mg/kg. The
 contaminated plume covered an  area of about 180,000 m2, and
 about  820,000 m3  of groundwater was  estimated  to be con-
 taminated. In the contaminated aquifer, anaerobic reducing condi-
 tions were  prevalent  with high  iron  concentrations (140 mg
 Fe2+/l), negative Eh and low pH-values (3.1-7). No arsenic was
 found  in the River Rhein, only 300 miles north of the  sludge
 disposal site.
  Four possible remedial action alternatives were discussed:
 •Excavation and treatment of polluted soil
 •Extraction and  treatment of contaminated groundwater
^ncapsulatioiLQf the heavily contaminated area
^Jn3lu_oxidation)of arsenic compounds
  The fourth alternative was chosen.  Four years of groundwater
 monitoring indicated that the plume was shrinking and the Eh and
 pH-values increasing (Fig. 1). It therefore appeared possible to ox-
 idize the trivalent  arsenic  into  pentavalent  arsenic  and cause
 precipitation of complex  arsenic-iron-manganese  compounds.
Laboratory tests with bleach (NaOCl), hydrogen peroxide (H2O2)
and potassium permanganate (KMnO4) led to the decision to use a
solution of 2 g KMnO4/l for injection into the ground. 29,000 kg
of KMnO (0.472 g/m3 of water saturated sediment) were injected
into 17 wells and piezometer wells. Arsenic concentrations were
reduced in average from 13.6 mg/1 in 1975 to 0.06 mg/1 in 1977.
However, an increase to 0.4 mg/1 in 1979 indicated that the mixing
of contaminated water and oxidizing solution was not sufficient.
  Total costs for disposal of arsenic containing sludges, monitoring
and injection of KMnO4 amounted to DM 750,000.1A8 [Ed. note: 1
DM =  $0.34 US]

TETRACHLOROETHYLENE REMOVAL
FROM GROUNDWATER
  In 1979, a storage  tank in Sindelfingen, Land Baden-Wuerttem-
berg was overfilled  with tetrachloroethylene, spilling and con-
taminating the soil and groundwater. The contaminated soil was
immediately excavated and  disposed of as  a hazardous waste. A
mixture of dissolved  and liquid chlorinated hydrocarbons was then
pumped from the bottom of the  excavated pit, and the polluted
groundwater was treated in a mobile activated carbon filter unit
and discharged into a sewer. Investigation into the migration of the
hydrocarbons revealed high concentrations  of tetrachloroethylene
in the aquifer. An estimated 14 metric tons of hydrocarbons had
been discharged into  the aquifer and had moved upstream.
      RHINE RIVER
                  1971 	
                  1975	
                  ISO-CONCENTRATION LINES
                                              150m
CONTAMINATED SITE
                        Figure 1
        Shrinking of the Plume of Arsenic in Groundwater
                                                                            INTERNATIONAL ACTIVITIES
                                                    565

-------
  An attempt was made to leach the chlorinated hydrocarbons in
the 2,500 m2 spill area.  Water was infiltrated into the ground in
90 m long ditches at the rate of 2 mVhr. Within 18 months, pollu-
tion had decreased by  50%.  The leaching  liquid  and polluted
groundwater were pumped out of eight wells and treated in an ac-
tivated carbon  filter. The  treated water  was used  for further
leaching, and within 18  months, 17 metric tons of hydrocarbons
were recovered.li2'3

IN SITU TREATMENT OF GROUNDWATER
CONTAMINATED WITH HYDROCARBONS

  Groundwater  contamination  with  aromatic and  aliphatic
hydrocarbons was discovered on an industrial site near Franken-
thal,  Land  Rheinland  Pfalz. This  contamination  was caused
primarily by fuel oil along with benzene, xylene, toluene, naptha-
lene and styrene seeping  into the ground. While liquid  fuel oil was
pumped  out, remaining hydrocarbons saturating the soil were
estimated at 20 to 30 metric tons. Aromatic hydrocarbons were
found downstream of the site in the direction of a municipal water
works. The plume migration  had stopped, however, due to  pump-
ing of the shallow aquifer.
  The upper Rhine Valley aquifer is sedimentary, mainly from the
quaternary period. It has an irregularly layered structure with sand
and gravel  layers of variable thickness and horizontal extension
separated by clay and loam lenses or aquitards. In the area of the
contaminated site, a distinct high permeable layer of approximately
16 km thickness  (Kf  = 5.10~4 m/sec) overlies a clay barrier of a
thickness of 48 km or more. The aquifer has a microstructure rang-
ing from fine sand to coarse  gravelly sand.
  After three years of pumping, the water table in the  aquifer had
been lowered to a 1 m depth above the  impermeable layer. Ground-
water flow was extremely low.
  Combined hydraulic flushing and induced biodegradation of the
hydrocarbons was favored over containment  of the contaminated
aquifer by using  impermeable walls.' It was decided to take a two-
step remedial action:
•Hydraulic measures to control the flow of groundwater
•Biological and chemical treatment of contaminated groundwater
 in situ and in an on-site treatment plant
  The responsible water authority required that nutrients injected
into the aquifer to accelerate biodegradation and the flushed con-
taminants be kept  within a defined area so that the surrounding
aquifer was not contaminated.
  The effects of the hydraulic  measures were simulated  on a
numerical groundwater flow model. The model was calibrated at
the first stage of the action when the aquifer had a level of 1  to 6 m
of clean groundwater (drinking water standard) at a steady state
flow with all polluted parts of the aquifer "under water".
  Two separate  recirculation  lines were installed,  one for  the
flushing water (5  I/sec) and the second for clean injection water (20
to 30 I/sec). The latter was operated  throughout the remedial ac-
tion to control spreading of  contaminants  from the treated area.
The recirculated  flushing water, contaminated with hydrocarbons
and biodegradation by-products, was  stripped and filtered  before
re-infiltration (Fig. 2). Biodegradation  was enhanced by controlling
the dosage of the nutrient  nitrate and  by increasing the  water
temperature 10°C. Laboratory experiments revealed that microbes
present in the soil (  5,000 microbes/g of soil) would degrade
gasoline and benzene. Biodegradation of aromatic hydrocarbons
was simpler than  degradation of aliphatic  hydrocarbons,  and
benzene biodegradation was better than that of xylene and toluene
(Fig. 3). Microbial activity was controlled in the field  by measuring
nitrate concentration in the recirculated water. The maximum con-
centration was 500 mg/1, with an average of 300 to 400 mg/1.
  The effectiveness of the remedial measures was monitored using
samples from randomly  located monitoring wells over 4 months.
Significant differences in hydrocarbons concentrations were found.
After three months it was found that aromatics had been degraded
in the whole area, and aliphatics were reduced to about one-third of
their initial concentration.'•'•7'1
                                                                                     ±"^
                                                                                     ~^?™r	8L___
                                                                                 Figure 2
                                                                       Groundwater Treatment Scheme
                                                                                  ' '."i OF OPERATION  (MONTH)
                                                                                 Figure 3
                                                                       Biodegradation of Hydrocarbons
                                                       ENCAPSULATION OF HEXACHLOROCYCLOHEXANE
                                                       ON AN INDUSTRIAL SITE

                                                         In January, 1979, it was discovered that hexachlorocyclohexane
                                                       (HCH) had been disposed of on an industrial site in Gendorf, near
                                                       Munich,  Land Bayern. Residues from  the production  of the
                                                       pesticide LINDANE were the source of this waste.  The residues
                                                       were a mixture of various isomers of HCH and some other "im-
                                                       purities."
                                                         The HCH waste had been disposed of on a concrete slab of an
                                                       old building foundation and was covered with only 0.5 m of soil at
                                                       the  time  of initial investigation. Detailed  investigations  by the
                                                       Bavarian EPA revealed an HCH-waste layer of 2.0 to 2.5 m thick,
                                                       covering an area about 250 m2, 20 m above the groundwater table.
                                                       The waste contained part dry powder and part pasty sludge. Con-
                                                       tamination in the vicinity of the disposal site was very low: HCH
                                                       levels in milk were 1/10 to 1/100, and in water 1/1000 of the max-
                                                       imum  acceptable level.
                                                         Four alternatives for remedial action were considered:
                                                       •Removal and off-site disposal in a hazardous waste  landfill
                                                       •Removal and disposal in the Herfa-Neurode salt mine
                                                       •Removal and incineration
                                                       •Encapsulation
                                                         The three removal alternatives were dismissed because  of the
                                                       enormous  operational safety and emergency  procedures  which
                                                       would be necessary for an excavation. In addition, the potential by-
                                                       products of incineration of the HCH waste included 0.7 metric tons
                                                       of chloride per metric ton of waste as well as dioxin or phosgene.
                                                       Taking "no action" would also be unsatisfactory, as infiltrating
566
INTERNATIONAL ACTIVITIES

-------
rainwater could cause HCH to leach into the groundwater, and
HCH could also volatilize and escape into the atmosphere.
  The site was therefore encapsulted to prevent these  potential
discharges from occurring. The vertical barrier, made of steel sheet
piles (LARSEN profile No. 20), goes down to a depth of 3 m and is
426 m in length. The site is covered with an impermeable cap made
up of 20 cm compacted gravel, 5 cm compacted sand, a 3 mm poly-
vinylethylene protective layer and a 1.5 mm PVC membrane. This
cap overlaps the vertical barrier and is buried to a depth of 60 cm.
A 60 cm layer of sand protects the  PVC membrane and is covered
by 20 cm of gravel, 8 cm asphalt and 3 cm asphalt concrete. The
final cover has a 3% slope.
  The site will receive long-term monitoring at three wells and has
been designated in the register of land  property to  prevent  any
damage to the encapsulation in the future. The total cost for encap-
sulation and monitoring was DM 250,000.

DIOXIN-CONTAINING WASTE  IN A
HAZARDOUS WASTE LANDFILL

  It was learned in 1983 that between 1969 and 1972, approximate-
ly 370 metric tons of dioxin containing waste (200 drums) had been
disposed of in the Gerolsheim hazardous waste landfill. This dioxin
waste was the result of the manufacturing of 2, 4, 5-T. Since 1972,
the barrels containing dioxin had been covered possibly by acid tar
wastes and by 20 m of unknown hazardous waste.
  Adjacent to Gerolsheim lies Hessheim, a solid waste dump site.
Both sites, still in operation, are located in former sand pits,  and
neither has a liner or leachate  collection system. The Gerolsheim
site is about 15 hectares and the Hersheim, 17 hectares. Gerolsheim
has received approximately 3.2 million metric tons of waste, about
20% of which is solid waste.
  The information about the  dioxin  wastes  was released by the
generator, the Christopher Boehringer Company, at  a time when
there was high interest in the burial location of dioxin  wastes in
Europe, particularly the 41 Seveso  drums.
  The responsible authorities initiated immediate investigations to
determine if there were any dioxin emissions. Dioxin was not found
in groundwater, nor in the soil, nor in plants or animals. The next
steps were to: (1) pinpoint the specific location of the dioxin waste
in the landfill; (2) evaluate the likelihood of its migration; (3) con-
duct a risk assessment; and (4) propose options for remedial action.
   Results of geological investigations  are shown in Figure 4.  The
 second aquifer was  found to  be contaminated, leading to the
                    monitoring wells  	
                    vertical barrier  	
                    discharge  wells  \	
                    monitoring wells  J

                    expected leachate table
   to water treatment
       3rd aquife
               leachate flow into
               2nd aquifer
               trough ..windows"
tertiary silt
quaternary silt
                           Figure 4
         Remedial Action Proposal for the Hazardous Waste
                    Landfill Site Gerolsheim
                          assumption that the quaternary sediments (KflQ-7 to 10 ~8 m/sec)
                          contained  "windows."  The third aquifer  was uncontaminated.
                          Leachates  from the landfill site which were polluting the ground-
                          water were primarily inorganic salts. Volatile chlorinated hydrocar-
                          bon  contamination downstream from the  site was significantly
                          higher than upstream (up to 729 /*g/l). Analysis  of downstream
                          sediments showed hydrocarbon contamination to a depth of 36 m.
                          These sediments contained 100 mg volatile and non-volatile hydro-
                          carbons/kg along  with  phenols and other coupling substances.
                          Neither dioxin nor heavy metals were found in these extensive in-
                          vestigations.
                            Gases being emitted from the site include methane, hydrogen
                          sulfide, halogenated hydrocarbons and aromatic compounds.
                            The emotions of citizen groups from the villages of Gerolsheim
                          and Hessheim have influenced discussions on remedial action. The
                          citizens have demanded removal of all of the wastes, but in par-
                          ticular the dioxin waste. Authorities consider this impossible and
                          unnecessary because of the dangers of excavation, the lack of
                          available disposal facilities  for dioxin wastes and because of the
                          relative lack of groundwater contamination. Encapsulation appears
                          to be the appropriate remedial response as  it will protect against
                          potential impacts of the waste, including groundwater contamina-
                          tion and air pollution. Encapsulation proposals include construc-
                          tion of vertical barriers extending below the second aquifer (30 to
                          50 m),  and placement of a nearly impermeable (not greater than
                          10-9 m/sec) cap. The expected leachate 3.8mVsec)  will either be
                          treated in an industrial sewage treatment plant or in a separate on-
                          site treatment facility.
                            The groundwater will be  tested for pH, electrical  conductivity,
                          dissolved solids, chlorides, sulphur peroxide, chemical oxygen  de-
                          mand, total organic  carbon, volatile halogenated hydrocarbons,
                          total cyanides, free cyanides,  phenol index, phenol index after
                          distillation, heavy  metals (mercury, cadmium,  lead, chromium,
                          nickel,  copper and zine) and toxicity (Beckmann rapid  analysis
                          Microtox). If the phenol index after distillation is greater than  100
                          /tg/1, groundwater will be analyzed with gas chromatography and
                          mass spectrocopy for 2, 4, 5-trichlorophenol. If the result is greater
                          than 100 us/I, a 2, 3, 7, 8-TCDD analysis will be necessary.
CONCLUSIONS
  While only a few case studies have been illustrated, they repre-
sent trends of remedial activities in the Federal Republic of Ger-
many. Hydraulic measures combined with groundwater treatment
are used extensively to cope with  contamination  by halogenated
hydrocarbons,3  and enhancement  of  microbial  degradation  in
aquifers is a promising approach for the future.
  Although small hazardous waste dump sites were excavated and
the waste disposed of according to  present standards in hazardous
waste disposal facilities, encapsulation is presented as the only ap-
propriate remedial alternative, particularly for larger dump  sites.
This  is  particularly  true  where  dioxin-containing  wastes are
suspected, because of the environmental risks  connected with ex-
cavation and  the problems surrounding its disposal in licensed
disposal  facilities. While  encapsulation, using vertical barriers,
slurry walls and clay caps or membranes may allow time to explore
more  permanent and effective solutions, it is likely that  perpetual
maintenance and reconstruction of the "container" will occur in-
stead.
  In situ treatment of contaminants in soil or  groundwater is not
possible if there are heterogeneously distributed mixtures of con-
taminants. Only in very rare cases, as with the arsenic  treatment
described, does it offer cost-effective solutions.
  From these remedial actions at abandoned hazardous waste sites
and contaminated industrial sites,  it becomes obvious that careful
handling and  storage of hazardous substances and  treatment of
hazardous wastes  in appropriate  facilities to  render it harmless
before landfilling, will ultimately be much more cost-effective than
the best available remedial action.
                                                                                     INTERNATIONAL ACTIVITIES
                                                                                    567

-------
 REFERENCES

  1. Schoettler, U., "Treatment of Contaminated Ground water as Reme-
    dial  Measure" Insitu  fuer  Wasserforschung  GmbH,  Dortmund
    Februar 1984 (Fellowship report for NATO/CCMS Pilot Study "On
    Contaminated Land").
  2. Informationsschrift  der  Laenderarbeitsgemeinschaft  Abfall Nr.  5
    "Gefaehrdungsabschaetzung und Sanierungsmoeglichkeiten bei Alta-
    blagerungen" Erich Schmidt Verlag, Berlin,  1982.
  3. Anonymous: "Leitfaden fuer die Beurteilung und Behandlung  von
    Grundwasserverunreinigungen  durch  leichtfluchtige  Chlorkohlen-
    wasserstoffe" Ministerium fuer Ernaehrung, Landwirtschaft, Umwelt
    and  Forsten des Landes Baden-Wuerttemberg, Stuttgart 1983.
  4. Harres, H.  und  Holzwarth,  W.:  "Sanierungsmoeglichkeiten  bei
    Boden- und  Grundwasserverunreinigungen  mit leichtfleuchtigen
    Chlorkohlenwasserstoffen" Z. dt. Geol. Ges., 134, 821-831,  1983.
  5. Geldner, P.: "Anwendung mathematischer  Modelle im Zusammen-
    hang mit Grundwassersanierungcn im Bereich von  Altablagerungen"
    in "Sanierung Kontaminierter Standorte  - Dokumcnialion eines
    Arbeilsgespraches im April  1983"  BMFT/Umweltbundesamt,  Ber-
    lin,  1983.
  6. Geldner, P.: "Removal of Hydrocarbons by Subsurface Biodegrada-
    tion—An  Engineering  Application—"Colloquium"  Ontwikkeling
                                                               bodemreinigingstechniken"  5. April 1984 in Ede, Netherland,  Di-
                                                               recloraat Generaal voor de Milicuhygiene, Den Haag.
                                                            7.  Battermann, G.: "A Large-Scale Experiment of In Situ Biodegrada-
                                                               tion of Hydrocarbons in the Subsurface." Proc. Put. Symp. "Ground-
                                                               water in Water resources  Planning," by UNESCO-FAH-IAHS, in
                                                               Koblenz, FRO, Aug.. 1983.
                                                            8.  Matthes, G.:  "In  Situ Treatment of Arsenic Contaminated Ground-
                                                               water"  The Science of the Total Environment, 21 1981, 99-104, and
                                                               in Quality of Croundwater, Proc. of an Intern. Symposium, Noord-
                                                               wijkerhout. The Netherlands, March 1981,  W. van Duijvenbooden,
                                                               P. Glasbergen and H. van Lelyveld (Eds.),  Studies in Env. Science,
                                                               17, Elsevier Scientific Publish. Comp.
                                                            9.  Defregger,  F.:  "Management of Uncontrolled  Hazardous  Waste
                                                               Sites in  Bavaria Illustrated by Closing an Industrial Chemical Waste
                                                               Dump," OECD Seminar  on Hazardous  Waste  "Problem" Sites,
                                                               Paris, Nov. 1980,  ENV/WMP/ SO.Sem.8.
                                                           10.  Tabasaran, o. and Thomanetz,  Ł.: "Wesentliche Untersuchungser-
                                                               gebnisse in Kurzfassung und Sanierungsvorschlage fur die Sonderab-
                                                               falldeponie Gerolsheim,"  Proc. FGU-Seminar,  Berlin  10-11 Mai
                                                               1984 "Dioxine, eine Gefahr fur jedermann?"
568
INTERNATIONAL ACTIVITIES

-------
  REVIEW OF THE  DEVELOPMENT  OF  REMEDIAL  ACTION
                TECHNIQUES  FOR SOIL  CONTAMINATION
                                   IN  THE NETHERLANDS

                                           DICK HOOGENDOORN
                           Laboratory  for Waste Matters and Emission Research
                     National Institute for Public Health and  Environmental Hygiene
                                          Bilthoven, The Netherlands
INTRODUCTION

  With the discovery in 1978 of severe soil contamination under a
new housing development in Lekkerkerk, the problem of soil con-
tamination was introduced to The Netherlands. Since that time,
an unflagging flood of contaminated sites has been discovered.
  At present,  it is estimated that the number  of contaminated
sites exceeds the 4000 uncovered during a 1980 inventory. About
1000 of these sites contain such severe contamination that some
kind of remedial action will be required.
  The large number of sites has demanded a stepped-up remedial
action program which  probably will not be finished before 1998.
A further increase of the number of contaminated sites might push
this date further forward.
  The remedial action activities in Lekkerkerk consisted of the
excavation of 100,000  m3  of contaminated soil and treatment of
the contaminated groundwater. The  soil was transported  to  a
domestic refuse incinerator. There it was stored and gradually in-
cinerated along with domestic waste. The total cost of the remedial
action program in Lekkerkerk amounted to about $65,000,000
(US).
  With a total yearly amount available for remedial action of about
$60,000,000 (US) for the next three years, it is evident that remed-
ial action like Lekkerkerk  is out of the question. Thus the devel-
opment of a more  cost-effective  remedial action technology is
necessary. In this paper, the author gives an overview of the poten-
tial application of remedial action techniques developed and being
developed in The Netherlands.

THE DUTCH SITUATION AND POLICY
  The Dutch situation involving  contaminated sites has some
unique  characteristics.  First, the population density is  high  and
therefore there  is a great need for land. A policy which accepts a
considerable amount of permanently contaminated land  unsuited
for certain types  of human use is, in  general, not acceptable.
Second, in many parts of the Netherlands, particularly in the in-
dustrial areas where most  of the contaminated  land is  situated,
the groundwater level is high. Consequently, the contaminated soil
is partially located in the saturated groundwater zone resulting in a
serious  danger of further  propagation of the contamination via
groundwater flow. Since mud groundwater is a source of drinking
water, this situation is highly undesirable.
  Finally, the character of the soil requires considerable  effort in
preparing a site for building. In the past decades, there has been
extensive development of  housing. For site preparation, demo-
lition rubbish was often used as a cheap fill material.  It seems
that often, wittingly or unwittingly, great amounts of hazardous
wastes were included with the harmless rubbish. This resulted in a
number  of new  housing developments  being  constructed  on
seriously contaminated soil. Lekkerkerk is the most well known but
certainly not the most severe example of these sites.
  Because of this and similar situations, a considerable number
of cases required quick remedial action. Since the required tech-
nology was not yet available, this cleanup program resulted in the
construction of temporary storage facilities, in  which contam-
inated soil is stored while awaiting ultimate disposal.
  The situation in The Netherlands resulted in a soil contamina-
tion policy based on the  principle that the remedial action pro-
gram must remove the contaminants from the soil. Containment is
only acceptable when the removal would result in disproportionate
cost or would cause an unacceptable risk to public health.  More-
over,  the  excavation  and transportation of contaminated soil
abroad is unwanted. Project proposals containing such solutions
are not funded under the framework of the Soil Clean Up (Interim)
Act.
  Because of this policy, this paper only deals with the technology
concerned with in situ soil treatment and soil treatment after ex-
cavation.

CLASSIFICATION OF REMEDIAL
ACTION TECHNIQUES
  This paper described technology which is now operational in the
Netherlands or which is presently being developed and is expected
to become operational within a reasonable time. A more detailed
description of the Dutch situation is given in Reference 1. A more
comprehensive review which also contains more theoretical tech-
niques is given in References  2 and 3. The remedial action tech-
niques can be subdivided into three main categories (Fig. 1):
•Thermal treatment
•Treatment by extraction
•Biological treatment
Thermal Treatment

  The contamination is removed by heating the contaminated soil.
First, the soil is excavated. Then thermal treatment is performed in
an installation with direct (convection or radiation) or indirect
(conduction) heat transfer.
  When the temperature in the installation is relatively low, incom-
plete incineration of the contaminants occurs. Post-treatment of
the gas must then be  effected. The required destruction can be
achieved by incineration at high temperatures in an afterburner,
thermal treatment at moderate temperatures using appropriate cat-
lysts or treatment at low temperatures followed by scrubbing of the
gas and purification of the scrubbing liquid.
                                                                             INTERNATIONAL ACTIVITIES
                                                     569

-------
                    Treatment after excavation
Thermal treatment
Extraction
Microbiological
  treatment
                                       steam stripping
                        evaporation by thermal treatment
                                       (T=± 300°C)
                        evaporation by thermal treatment
                                       (T=± 700°C)
                              incineration (T > 800 °C)

                                     aqueous solution
                                       organic solvent
                                            flotation

                                         landfarming
                                          composting

                                    (industrial system)
                        In situ treatment
Thermal treatment

Extraction
Microbiological
  treatment
                                       steam stripping

                                     aqueous solution

                                         landfarming

                                        bioextraction
                           Figure 1
              Summary of Remedial Action Techniques
   In the case of in situ treatment, steam is injected into the soil.
The contaminants diffuse into the gas phase and then are trans-
ferred to the surface where, if necessary, adequate post-treatment
takes place.

Extraction

   The extraction process consists basically of mixing the contam-
inated soil with an extracting agent to transfer the contaminants
from the soil particles to the extracting agent. With this process, it
is possible to remove not only contaminants that are soluble in the
extracting agent (in general an aqueous solution, but sometimes an
organic solvent), but also contaminants that are in fact insoluble in
the extracting agent. In the latter case, it is required that the ex-
tracting agent have properties favorable to the formation of stable
colloidal suspensions of the contaminants.
   After extraction, the purified soil particles are separated from
the extracting agent containing the contaminants. The extracting
fluid is subsequently purified.

Microbiological Treatment

   Microbiological  treatment implies the removal of contaminants
by biodegradation. After excavation of the soil, this treatment can
be accomplished by landfarming, composting or treatment in a de-
signed installation. The last possibility is still more or less theo-
retical  since  it  is  dubious  that  a  sufficiently  high biogradation
velocity can  be reached to result in a sufficiently low  residence
time.
   In situ microbiological treatment can take place  by landfarm-
ing and bioextraction. The former technique is only applicable
when the contaminants are concentrated in a top soil layer having
a maximum depth of about 0.5 m.

POTENTIAL APPLICATION OF
REMEDIAL ACTION TECHNIQUES

  The potential for successful application of the  various remedial
action techniques depends on  the type of contaminants, the type of
soil and the location of the contaminants with regard to the ground
level. These three aspects are discussed below.
 Type of Contaminant

   Contaminants can be subdivided into organic and inorganic. The
 inorganic category can be surther subdivided into heavy metals and
 metalloids on one hand and cyanides and cyanide complexes on the
 other hand. Finally, a miscellaneous category  of inorganic con-
 taminants consisting of acids, phosphates and ammonia remains.
   The organic contaminants are subdivided into aliphatic and aro-
 matic  hydrocarbons,  polynuclear  hydrocarbons,  halogenated
 hydrocarbons  and pesticides. These  subdivisions finally result in
 seven main categories of contaminants:
 •Aliphatic and aromatic hydrocarbons
 •Polynuclear hydrocarbons
 •Halogenated hydrocarbons
 •Pesticides
j • Heavy metals and metalloids
I 'Cyanides and cyanide complexes
I •Miscellaneous (acids, phosphates, ammonia, etc.)
 Type of Soil

   In Reference 1, a subdivision of types of soil into five categories
 was made:
 •Sandy soils
 •Loamy soils
 •Clay
 •Peat
 •Stratified soils
   For each of these soils, the application of each of the remedial
 action techniques can be determined. However, the present state-
 of-the-art of the remedial action techniques makes such a detailed
 differentiation by type of soil premature.
   Therefore, in practice, a differentiation between only two types
 of soil is made. There are the sandy and loamy soils which can be
 cleansed relatively easily, and there are the clay, peat and stratified
 soils where the application of remedial  action techniques is lim-
 ited. In this paper, the latter differentiation is used.
                                                        Location of Contaminants

                                                          The location of the contaminants with regard to the ground level
                                                        is important for two reasons:  (1) the potential for excavation of
                                                        contaminated soils is practically limited and (2) certain in situ re-
                                                        medial action techniques can be applied only when  the contam-
                                                        inants are located near the ground level  (landfarming and steam
                                                        stripping).
                                                        Remedial Action Evaluation

                                                          The  evaluation of techniques based on  type of contaminants
                                                        and type of soils has been accomplished, and the results are given
                                                        in Tables  1 through 4. Tables 1 and 2 deal with remedial action
                                                        techniques after excavation, while Tables 3  and 4 deal with in situ
                                                        treatment.  The tables are  mainly  based on  information derived
                                                        from Reference 1.
                                                          The symbols used in the tables are:
                                                        +     applicable
                                                        + /- applicable in some cases
                                                        - / + in general, not applicable
                                                        -     not applicable
                                                          It must be emphasized that these tables describe potential appli-
                                                        cations of techniques, not actual ones.
                                                          Given a specific contaminant and a type  of soil, one can deter-
                                                        mine which technique or techniques can be applied by using Tables
                                                        1  through 4. In case more than one technique is applicable, a choice
                                                        must be made.
                                                          At present, the actual availability of a technique often deter-
                                                        mines the final choice. However, in the near future, when the total
                                                        number of remedial possibilities will certainly increase, criteria
                                                        other than availability will become important. In Tables 5 and 6,
570
INTERNATIONAL ACTIVITIES

-------
                           Table 1
     Potential Applicability of Remedial Action Techniques After
               Excavation; Sandy and Loamy Soils
                           Table 2
     Potential Applicability of Remedial Action Techniques After
             Excavation; Clay, Peat and Stratified Soils

aliphatic and
aromatic
hydrocarbons

P y y

low boiling
point
high
boiling
point


halogenated hydrocarbons
pesticides
heavy neta Is a ad
nets Holds


• Iscallert




cyanide
complexes
eous
L IUT-H.I 1 t ri'.ifnciu
n.
3
i

-f


+ /-

-

-
-i <
•5
*i

4


+/-
+/-
-



I'i
O n
°.S'

4


+
4
+/-
-
4
4
-
-*i
3
n o

*


4
4
4

4
4

irv
by
:,
1
S
0
3
»/-


+/-
+/-
+/-
+/-
+/-
+/-
«/-
itm.MU

O
3
O
3
t/-


+ /-
+ /-
+ /-



cT
3
*/-


«/-
+ /-
*/-
+ /-
*/-
+ /-
+ /-
nl.-rn-b

1
3
a
r.
+
+
+/-
+ /-
+/-

+
-/+
Iol..Blcjl
|
tr.
3
4
4
+ /-
+ /-
+ /-

4
-/+


•iroiaatlc



polynuclea


point
high
bollint,
point
r hydrocarbons
halogena ted hydrocarbons
pest Icldes
heavy netals and
•etallo'tds


mi sea 1 lenn




cyanide
complexes
eous


p
-i
I'
+ /-


-
-
-


H
0
= =
4
*/-
+ /-
+ /-
-
-


It "?
0 "
°.o
a
r-.
4
4
4
+
+ /-

+
-
=
= -
— =
4
+
+
+
+

4
4




•±

5

-
-
-

+/-

-/+

-/+
-/+
-/+
-/+
-/+
-/+


3
-
-
-
-
-
-/+





3
3
3
r.
+ /-
+ /-
+ /-
-/+


4
-/+
fotri,;!,-.) |
11
3
V.
3
r.
+/-
+/-
+/-
-/+


+
-/+
some of these criteria are given for techniques to be used after ex-
cavation  and for in situ techniques.  The symbols used in  these
tables are as follows:
 +     favorable or not problematic
 0     limiting or slightly negative
       strongly limiting or negative

  The following paragraphs contain explanatory notes on many of
the criteria.
•Energy  demand is extremely important for the thermal  tech-
 niques.  With increasing temperatures in the process unit, energy
 costs increase; hence, increasing temperature is a negative feature.
•Treatment by extraction and/or  microbiological treatment  often
 requires the addition of other chemicals such as nutrients and/or
 oxygen.  Techniques are judged favorable if these additions can be
 avoided.
  The amount of residue from thermal techniques, if they are pro-
vided with adequate post-treatment of the gases, is negligible. The
extraction techniques are judged negative on this feature, especially
when the  percentage of fine particles in the soil increases.
  When more than  one contaminant is present, the limited  num-
ber of contaminants which can be removed  by steam stripping is
limited. Hence, the  technique is generally judged to be negative.
The  application  of  the remaining thermal  techniques  is limited
when the contaminated soil contains heavy  metals and/or metal-
loids. Treatment by extraction is very attractive in this case.
  Microbiological techniques normally degrade only a single chem-
ical.  Thus,  when numerous contaminants are present  simultan-
eously, microbiological  techniques are not highly rated.
  A  reliable estimate of  the cost  is difficult to make at present;
most of the techniques are still in development and only a few have
been used in the field.  As a result, only a range within which  the
real costs are expected  to vary is  given in Table 5. The amounts
reported include only the actual remedial action cost excluding the
cost for excavation, transport and possible cost for dumping the
cleaned soil afterward. The costs are strongly determined by the
type of soil.
  As discussed before, sandy and loamy soils are relatively easy to
cleanse. Therefore, the lower amount in the range given in Table 5
in general will be valid for these types of soils. Conversely, the cost
of cleaning up clay, peat and stratified soils will be near the upper
limit of the given range.
  Any required increase in temperature in the thermal techniques
will result in an increase in cost. Comparing thermal and extrac-
tion techniques  shows  that  steam  stripping  and  evaporation at
lower temperatures (± 300 ° C) are cheaper than extraction. Evap-
oration at higher temperatures (±  700° C) and incineration are
more expensive than  extraction. The cost of microbiological de-
gradation is promising on comparison with the other techniques.
However, results of full scale application of this technique is not
currently  available. This makes any  definitive conclusion about
cost advantages premature.  Since  the cost of in situ treatment
strongly depends on the total treatment time, no reliable estimate
of these costs can be given.
  In Tables 1 through 6, the potential combinations of  different
techniques together with some features are given.
  During the last four years, much research has been performed
in order to develop full-scale remedial action techniques. In  the
next few paragraphy,  the current state-of-the-art of both the treat-
ment after excavation and in situ treatment is given. The review is
limited to those techniques which  are developed at least to pilot-
plant scale.
  However, various contractors have designed treatment installa-
tions which could result in either pilot-plant  or full-scale installa-
tions at any time. Due to uncertainty about the time required to ex-
ecute these designs, no attention is given to them.
  Most of the techniques to be described have been developed with
financial support from  the Dutch government. A more  extensive
review of the actual situation is given in Reference 4.
                                                                                      INTERNATIONAL ACTIVITIES
                                                           571

-------
                           Table 3
     Potential Applicability of In Situ Remedial Action Techniques;
                     Sandy and Loamy Soils
                                                                                    Tables
                                                           Selection Criteria for Remedial Action Techniques After Excavation

• 1 1 ptl,l t t C
• row* 1 1 c
h yd roc *r bo
..d

low bol I Inn
point
high
bol I ln«
polru
polynuclctr hydrocarbon*






p*«i |c i4ft
metalloids





It end


cy»nIJ! r*« Idw

tfh*n »or» conmlnjnt* «r«
j>rc««-n( *l«ul(*n*ou*ly
.


I ht'i

t
5
"
-3
ol-
•
o

-


IV)



.. *

" ',
a
«/-
'
•

o/-


JM



,' 7

J. 1
-
'
•

e


X»
.„


3
; -
• '

•
•

O


6*0
it,-

, „
l\
i •

•
o


*


120
......


3

•3
•
o


4


no



-
r

4
O
-

*


120
• l r c t


I.
•*

•
O
•




1M>
1 ,1 „; . .1


i
-

4
O
«




160
                           Table 4
     Potential Applicability of In Situ Remedial Action Techniques;
                  Clay, Peat and Stratified Soils
                                                                     Note: I Guilder-S0.31 US
                                                                                   Tabled
                                                               Selection Criteria for In Silo Remedial Action Techniques









po 1 ynucle


pcct Ic Ide
heavy met
•etJiltoTd



•14C4 I Irnr









r hy
(Wlllnit
point
rocJtrbom



d


d*
****

















-





1




-



•/-


«/-









•/-
•/-
'

«/-
-

'
-
-/>
















-

AVAILABLE TECHNIQUES FOR
TREATMENT AFTER EXCAVATION

Thermal Treatment

Evaporation at ± 300° C

  A full scale thermal treatment plant has been developed by Eco-
techniek (Fig. 2).  An extensive description of the installation  is
given in Reference 5. The  contaminated soil is heated to 200-
300° C, and the gases are burned at approximately 800° C with
                                                                     AMOunt of c«« Ido
                                                                     Co«t (dutch |ulld*ri/a'>
                                                        the addition of extra air.  In order to minimize the total energy
                                                        need, an ingenious heat recirculation system is used.
                                                          The total capacity of the installation  strongly depends on the
                                                        moisture content of the soil and the type of soil. For sandy soils
                                                        (low moisture content),  the capacity is  about 30 mVhr (48,000
                                                        mVyr) while for clay (high moisture content) this capacity reduces
                                                        to about 15 mVhr (24,000 mVyr).  In the latter case, the total load
                                                        of the installation still amounts 30 mVhr, but the need to mix the
                                                        feed with clean sandy soil  to achieve the optimum mixture in the
572
INTERNATIONAL ACTIVITIES

-------
installation reduces the soil cleaning capacity to 15 mVhr. The re-
duction in capacity can be avoided by mixing with  sandy soils
contaminated with substances which can be removed by a thermal
treatment system.
  Up to now, several  ten thousands of cubic meters of soil con-
taminated  with various  substances such  as  benzene, toluene,
xylene,  naphtalene, diesel oil  and gasoline have been successfully
treated using available incinerators.
Evaporation at ± 700° C

  The "Afvalverwerking Rijnmond" (AYR) in Rotterdam has an
installation for the incineration of domestic refuse. To a limited de-
gree, contaminated soil can be treated in this installation by mixing
it with  the domestic refuse.  Due  to the grate construction,  and
the mixing thereon, only a 10% soil-90% refuse mixture is possible.
This limits the capacity for contaminated soil to about 50,000 mV
hr.  The temperature in the installation exceeds 500° C while the
gases are heated up to at  least 850° C. The cleaned soil is mixed
with the slag from the domestic refuse, thus making reuse of the
soil questionable. The contaminated  soil  from Lekkerkerk  was
cleaned  (±100,000 m-n) in this unit. Although this installation
was not  developed  for the treatment  of  contaminated soil,  its
potential capacity requires mentioning it here.
  The Nederlandse  Beton Maatschappij (NBM) has developed a
pilot-scale incinerator consisting of a rotary kiln in which the con-
taminated soil is headed by a heat  exchanger depending  on the
type of  contaminants,  up  to a maximum temperature  of 850° C.
(Fig. 3). The gases are  incinerated  in an afterburner at a tempera-
ture of about 1,200 ° C.
  The capacity of the pilot-plant is 0.3 mVhr. When the test pro-
gram is  completed, scaling up to a full-scale plant with a capacity
of about 10  mVhr  (16,000 mVyr) is  foreseen.  The total  energy
demand is reduced by using heat recovery.
  The installation has  been tested with clean soil. The results in-
dicated that the design criteria with respect to the temperatures in
the rotary kiln and the afterburner were achieved. At the time of
the preparation of this  paper, no results of the tests with soil con-
taminated with cyanides were available.

Incineration at a Temperature  > 850 ° C
  The AYR disposes of wastes in  a hazardous wastes incinerator.
The installation consists of a rotary kiln in which a temperature of
about 1,300° C is maintained. It is estimated that an excess capacity
of about 5,000 mVyr is available.  This can be used for the incin-
eration  of  very severely  contaminated soils (containing  PCBs,
pesticides). Since this installation is not specially designed for the
incineration of contaminated  soil, no further attention has  been
paid to it here.
  Bob Kalis/Esmil  is  developing  an incinerator consisting  of a
fluidized bed in which  contaminated soil is incinerated at temper-
atures up to  850° C. At present, a pilot plant  with a capacity of
about 0.3 mVhr is available. When  test results are satisfactory, a
full-scale installation is planned with a capacity of about 17,000
mVyr. A process schematic of the installation is given in Figure 4.
Treatment by Extraction

Aqueous Solution
  The Hollandsche  Beton Groep NY (HBG) has developed,  with
the  support of the Netherlands Organization for Applied Scientific
Research (TNO), a full-scale extraction installation with a capacity
of about 15 mVhr (24,000 mVyr). During the test program  in a
pilot-plant, soil contaminated with cyanides, both free and com-
plexed, was cleaned successfully. Because of the flexible nature of
the  extraction process, it may be expected that  final applications
of this technique will  be  broader than simple  cyanide removal.
They may include, for example, heavy metals and metalloids and
polynuclear hydrocarbons. The present installation is limited to the
cleaning of sandy soils. Further research might result in an exten-
sion to soils with a higher percentage of fine particles (loamy soil).
The process scheme is shown in Figure 5.
         CONTAMINATED
         SOIL
                                FUEL     PREHEATED
                                        AIR
                           Figure 2
Treatment of Soil by Evaporation at 200-300 ° C and Destruction of Gases
                at About 800 ° C in an Afterburner
                           Figure 3
   Thermal Treatment by Indirect Heat Transfer and Evaporation of
                        Contaminants
                                       (CHLIMLA1U) AIR
 IHRtlHI AUl)
 AIR
                      IPREIHEATED    WATER
           SOIL


           GASES


           WATER
                           Figure 4
        Thermal Treatment by Incineration in a Fluidized Bed
                                                                                      INTERNATIONAL ACTIVITIES       573

-------
   Ecotechniek BV has developed an installation based  on wash-
ing contaminated soil with heated water. The capacity of the in-
stallation amounts to about 70 mVhr (110,000 m'/yr). The pri-
mary application appears to be sandy soils contaminated with oil.
Due to the heating of the soil, the contaminants are released from
the sand particles. Recently, the installation was used for the clean-
ing of a sandy beach that was contaminated by oil from a shipping
disaster.
   Bodemsanering Nederland BV (BSN) has developed a mobile in-
stallation (Fig. 6) in which oil-like contaminants are broken loose
from soil particles by passing the contaminated soil through a water
jet curtain (water velocity: 220 M/sec). Clean soil particles are re-
moved from the resulting  slurry  by two hydrocyclones.  Contam-
inants are removed in an oil-water separator.  The water is recycled
to the water jet but has to be replaced daily. The capacity  of the
installation is about 7.5 mVhr (12,000 m'/yr).
   Although experiences with the  unit have been obtained through
cleaning contaminated soils at oil  refineries (and the cleanup  was to
the entire satisfaction of the scientists involved), the final concen-
tration of contaminants does not meet the criteria for cleaned soil
given by the Dutch government. Recently, redesign  of the installa-
tion to meet the soil criteria mentioned above was begun.

Flotation
   The Heidemij/Mosmans combination has developed a so-called
"froth flotation" process (Fig. 7). A pilot-plant installation  with a
capacity of about 3 mVhr has been built  and successfully utilized
with sandy soils contaminated with oil, including some polynuclear
hydrocarbons. The installation may also be suitable for treating
(sandy)  soils  contaminated with heavy  metals  and  metalloids,
cyanide compounds and chlorinated hydrocarbons. The separated
foam is transported to a specialized waste processor. The decision
to construct a full-scale installation with a capacity of about 25 m1/
hr (40,000 m Vyr) is expected shortly.
CONTAMINATED
SOIL
                          Figure 6
       Treatment of Soil by Extraction Through a Water Jet
                                            sou

                                            WATl H
                                                                      CONTAMINATED
                                                                      SOIL
                                                                                                                          WASTE
                                                                                                   CLEAN SOIL
                                                                     OF WATER
                           Figure 5
       Treatment of Soil by Extraction with an Aqueous Solution

574       INTERNATIONAL ACTIVITIES
                          Figure 7
                 Treatment of Soil by Flotation

-------
Microbiological Treatment

  The  results  of a feasibility study6  indicate that  many organic
contaminants in the soil are biodegradable when favorable con-
ditions are created. This general conclusion has convinced  the
Dutch  government to encourage further  research in this area.
Financial support will be given to research projects  designed to
determine optimal conditions for the application of landfarming
and  composting  techniques  to contaminated soils. In addition,
the solution to the particular problem  of a considerable amount of
soil  contaminated with HCHs is studied  by  the  "Rijksinstituut
voor Natuurbeheer (RIN)". The  preliminary results of this study
are quite promising.7 Anticipating these developments, some con-
tractors already offer  full-scale solutions for soils contaminated
with oil. Results of applications are not available at this moment.
 AVAILABLE TECHNIQUES FOR
 IN SITU TREATMENT

 Thermal Treatment

   In the summer of 1983, some tests were performed with the in
 situ injection of steam in the contaminated Griftpark in Utrecht.
 No final conclusions on the potential applications of this method
 can be drawn at this point. Therefore, some additional tests are
 planned on a sandy soil contaminated with volatile halogenated
 hydrocarbons. Testing will be limited to the upper 2 m of soil.

 Microbiological Treatment

   In  situ microbiological treatment  is in the  same situation as
 microbiological treatment after  excavation. The  reader is, there-
 fore,  referred to the relevant paragraph of this paper on that topic.
 In addition,  the government intends  to give financial  support to
 the further development of bioextraction techniques.


 CONCLUSIONS

   From the  preceding  paragraphs, it can be seen that the devel-
 opment of remedial  action techniques in The Netherlands is quite
 promising. Additional research, to extend the  range of potential
 applications  for the  techniques already developed to full-scale in-
 stallation, will be required. Especially for the clay, peat and strat-
 ified  soils  contaminated with  heavy metals and metalloids, no
 appropriate cleanup technique is available at this time.  Since re-
 medial action is urgent  while cleanup techniques will be available
 for most contaminants within a few years. The most  promising
 techniques are those used after excavation as opposed to those em-
 ployed in situ.
Treatment after
  excavation:
                                                                   Containment:
DEFINITIONS

  For a better understanding of this paper, some relevant defi-
nitions are listed below:
In situ treatment:            remedial action leading to the actual
                            removal of the contaminants without
                            any appreciable excavation of  the
                            soil.
                            cleanup  of the contaminated  soil
                            after it  has been excavated to  the
                            required degree; the remedial action
                            can take place both on the contam-
                            inated site or wherever the required
                            technology is available. In the latter
                            case, transportation of the contam-
                            inated soil is necessary.
                            remedial action resulting  in the pre-
                            vention of further propagation of the
                            contamination;  containment    can
                            take place both in situ (geohydrolog-
                            ical measures, encapsulation, barri-
                            ers, etc.) or after excavation (immo-
                            bilization,  waste disposal site, etc.).
                            storage of the contaminated soil
                            after  excavation  awaiting  further
                            remedial action; in fact,  temporary
                            storage is a specific example of con-
                            tainment.

REFERENCES

1.  "Market aspects of remedial measures", Jan. 1984.*
2.  Rulkens,  W.H.,  Assink, J.W. and van Gemert, W.J.Th., On-site
  processing of contaminated soil, Draft Report on Contaminated Land,
   NATO/CCMS Study Group on Contaminated Land, Building  Re-
   search Establishment, 1984.
3.  Banning, D.E., In situ treatment, Draft Report on Contaminated Land,
   NATO/CCMS Study Group on Contaminated Land, Building  Re-
   search Establishment, 1984.
4.  Handbook ofRemedial Action, Staatsuitgeverij, July 1983.*
5.  Reintjes,  R.C. and Schuler, C., The development of the  thermal soil
   cleaning installation, Ecotechniek B.V., Jan. 1983.*
6.  Hanstveit, A.O., van Gemert, W.J.Th., Janssen, D.B., Rulkens,
   W.H. and van Veen, H.J., Literature study on the feasibility of micro-
   biological decontamination of polluted soils, TNO, Mar. 1984.
7.  Doelman, P., "The cleaning capacity of soil with regard to HCH",
   Colloquium Development of  Remedial Achon Techniques, Ede, April
   1984.*
*In Dutch.
Temporary  storage:
                                                                                      INTERNATIONAL ACTIVITIES
                                                          575

-------
             EXTRACTION  AS A METHOD  FOR CLEANING
                   CONTAMINATED  SOIL:  POSSIBILITIES,
                              PROBLEMS AND  RESEARCH

                                           W.H. RULKENS, Ph.D.
                                                 J.W. ASSINK
                         Netherlands Organisation for  Applied Scientific Research
                                     Division of Technology for Society
                                         Apeldoorn, The Netherlands
INTRODUCTION
  The remedial methods used lo clean up contaminated soil can
be broadly divided into two categories:1'2-3'"• '• '•l6'M
•Those methods aimed at preventing or restricting the dispersion of
 the contamination to the immediate surroundings
•Those methods aimed at removing or destroying  the contamina-
 tion
  The remedial steps aimed at preventing or restricting the disper-
sion of the contamination to the surroundings include:
•Excavation of the  soil  and its transportation  to a permanent or
 temporary storage site. Temporary storage  of contaminated soil
 may be necessary if no suitable method  of treatment is available
 at the time, a method has to be developed or if the supply of soil
 exceeds existing treatment capacity.
•Restriction of the  dispersion of the contamination by installing
 vertical and  horizontal barriers  (bitumen membranes,  plastic
 sheets, bentonite-cement walls, steel sheeting, mastic layers, etc.).
•Restriction of the  dispersion of the  contamination by means of
 physical or chemical fixation techniques. The soil is treated with
 chemicals, such as lime, cement, water glass and  urea formalde-
 hyde, in such a way that the contaminants in the soil are immobi-
 lized to a greater or lesser extent.
•Restriction of the dispersion  of the  contamination by geo-
 hydrological measures (pumping off  groundwater  and infiltration
 of fresh water).
  The remedial methods aimed at removing or destroying the con-
tamination can be divided into two sub-groups:
•Excavation of the soil and cleaning it on  or  off-site. These
 methods are referred to as cleaning after excavation. The most im-
 portant are:
 •extraction
 "thermal treatment
 *steam stripping
 •chemical treatment
 •microbiological treatment
•On-site cleaning  of the  soil  without prior  excavation. These
 methods are usually referred to as in situ cleaning. The most im-
 portant techniques are:
 •extraction
 •steam stripping
 •chemical treatment
 •microbiological treatment
  The study described in this paper deals with the cleaning of ex-
cavated soil by means of extraction.  The following topics will be
discussed:
•A classification of contaminated soils into types
                                                   •A general description of the extraction process
                                                   •The equipment which can be used
                                                   •Field of application
                                                   •The present state-of-the-art
                                                   •The problem of residual materials
                                                   •The estimated costs of the cleaning process

                                                   CLASSIFICATION OF CONTAMINATED SOILS
                                                     Obviously, no  two cases of soil contamination are identical.
                                                   Differences in soil contamination cases include: soil type, contam-
                                                   inated  site size, site  location in relation to inhabited areas, the
                                                   depth to which the contamination has penetrated, soil permeabil-
                                                   ity, geohydrological situation, nature and concentration of the con-
                                                   taminants, concentration distribution of these contaminants and
                                                   potential danger to man and the environment.
                                                     To evaluate  extraction as a method of cleaning contaminated
                                                   soil,  the most  meaningful  classification of contaminated sites  is
                                                   based on type of soil and type of contamination.
                                                   Types of Soil

                                                     The different types of soil include:
                                                   •Sandy soils, which can  be subdivided into soils with a relatively
                                                   small  amount  of humus-like substances or clay particles and soils
                                                   with a relatively high amount of humus-like substances or clay
                                                   particles
                                                   •Loam and clay-like soils
                                                   •Peat and peat-like soils
                                                   •Soils of a highly heterogeneous composition, i.e., different types
                                                   of soils are present, usually in layers
                                                   •Dumps and other soils; in addition to the actual hazardous con-
                                                   tamination, large quantities of urban and/or non-hazardous in-
                                                   dustrial waste are present. This category' also includes contam-
                                                   inated soils under buildings and soils with contaminants in drums.

                                                   Types of Contaminants

                                                     The types of contaminants which can be encountered in soil can
                                                   be divided into  the following categories:
                                                   •Heavy metals  and metalloids (e.g., Cr, Co, Cu, Cd,  Ni, As, Zn,
                                                   Sn, Hg, Pb and Sb); these elements are  usually present as solid
                                                   compounds (e.g., oxides,  sulphates, sulphides, ferrites, nitrates,
                                                   halogenides, carbonates, silicates).
                                                   •Cyanides, both free (CN ~~) and complex (e.g., iron cyanides)
                                                   •Aliphatic and aromatic  hydrocarbons and  related substances
                                                   (e.g.,  mineral  oil, phenol, toluene, benzene, alcohol, monochlor-
                                                   inated hydrocarbons) and PCBs (polychlorinated biphenyls)
                                                   •Pesticides (e.g., lindane, aldrin, dieldrin)
576
INTERNATIONAL ACTIVITIES

-------
•Other components (e.g., ammonia, acids, lyes, phosphates, sul-
 phates and nitrates)
  This classification of contaminations corresponds broadly to the
classification  used in The Netherlands  in  the "Table of Test
Values" drawn up by the Dutch Ministry of Housing, Physical
Planning and Environment.7
Contaminant Format
  The above-mentioned contaminants can be present in  the soil in
widely differing forms:
•Present in the form of solid particles;  this category includes a
 large number of heavy metals and metalloids compounds.
•Present as a separate liquid phase, insoluble in water. The liquid
 phase can be present as drops or as a liquid film around the soil
 particles; contaminants of this type include mineral oil, petroleum
 and organic solvents (if they are present in high concentrations).
•Adsorbed physically or chemically as ions to the  soil particles
 (e.g., organic soil components and clay minerals);  heavy metals
 can belong to this category.
•Adsorbed as molecules to soil particles  (especially organic); this
 category  can include aliphatic and aromatic i compounds (if they
 are present inlow concentrations).
•Dissolved in the water phase between the soil particles.
GENERAL DESCRIPTION OF THE EXTRACTION PROCESS

  Basically, the extraction process consists of three steps (Fig. 1).
•The first step  consists  of the intensive mixing of the  extraction
 agent and the  contaminated soil.  The contaminants adsorbed  or
 attached  to the soil particles or located between the particles are
 dissolved or dispersed in the extraction agent.
•The second  step consists of the separation of the contaminated
 extraction agent and soil particles. Separation is usually combined
 with post-washing of the soil  with clean extraction agent in order
 to rinse out any contaminated extraction agent still present.
•The final step  consists  of cleaning the extraction agent. The con-
  taminants present are destroyed or separated as a residue, often in
 the form of a chemical waste. The extraction agent can then  be
 reused.
                                recirculation of purified extracting agent
                          Figure 1
     Extraction of Contaminated Soil (Simplified Process Scheme)
  Two principal mechanisms of removal can be distinguished in
extractive cleaning:
•The contaminants are dissolved in the extraction agent, with or
 without the aid of a chemical reaction preceding or simultaneous
 with the extraction.
•The contaminants are dispersed in the  extracting phase in the
 form of particles with or without the aid of prior mechanical treat-
 ment. The separation between contaminated  particles and soil
 particles can be based on differences in:
 *particle size (sieving)
 *sedimentation rate (classification)
 *surface properties (selective agglomeration/coagulation and flo-
  tation)
  In practice, combinations of the above can be used.
   A more detailed  diagram of the extraction  process which in-
 cludes prior and subsequent treatments is given in Figure 2. In this
 figure,  the following  successive  steps  can  be seen (the numbers
 correspond to those in Figure 2):
   1. The soil to be cleaned is pretreated to remove large objects
 such as pieces of wood,  plant remains, concrete,  stones, drums,
 etc., while hard clods of soil are reduced in size. The sieving residue
 can be cleaned separately.
   2. The pretreated soil is mixed intensively with an extraction
 agent. As already mentioned, the primary purpose of this step is to
 transfer the contaminants to the extraction fluid.
   3. The soil and the extraction agent are separated. In general,
 the contaminants, the smaller soil particles (clay and silt particles)
 and the soluble components in the soil are carried off with the ex-
 traction agent.
   4. The soil undergoes subsequent washing with  a clean extrac-
 tion agent to remove  as much remaining extraction fluid  as  pos-
 sible.
   5. The larger particles carried off with the extract phase are sep-
 arated as well as possible and, if necessary, undergo a subsequent
 washing with clean extraction agent.
   6. The contaminated extraction fluid is cleaned. Part of it is then
 reused, after the addition of chemicals if necessary.
   It is not always necessary to separate soil particles and the ex-
 traction agent before going on to the actual cleaning step for the
 extraction fluid. With certain types of contamination, the purifica-
 tion step can be applied directly to the suspension of soil par-
 ticles and extraction fluid. In that case, the separation of the soil
 particles from the extraction phase takes place after  the actual pur-
 ification step.

 Extraction Agents

   In general, an aqueous extraction agent is preferred. This pref-
 erence is based on a large number of considerations, such as:
 •Safety of the extracting agent for man and environment
 •Prevention of additional groundwater and air pollution
 •Natural presence of water in the soil
 •Purification possibilities of contaminated extracting agent
 •Ease of use
 •Costs of the extracting agent
   For the efficient operation of the extraction process, it will usual-
 ly be necessary  to add chemicals to the water and/or to heat it.
 Among the chemicals which can be added are:
 •Acids, such as HC1,  H2SC»4 and HNOs; the  primary purpose of
 these acids is to dissolve the impurities.
 •Bases, such as Na2CC>3  and NaOH; the purpose of these sub-
 stances is either to dissolve the impurities or to  disperse insoluble
 impurities in the extraction phase.
 •Surface active agents; addition of these agents aids dispersion.
 •Sequestering agents (complex formers) such as citric acid, ammon-
 ium acetate,  NTA and EDTA; these substances have a positive
 effect on the solubility of the impurities in the aqueous extract-
 ing agent.
   In addition to the separate use of the  above-mentioned chem-
icals, combinations of them can also be considered. For example,
acids and/or lyes may be used in combination  with complex-form-
ers. The extraction process can also  be favorably  influenced by
prior oxidation  of the contaminants  with the aid  of an oxidizer
 (e.g., hydrogen peroxide or ozone).
   In principle, it is also possible to employ organic  solvents as ex-
traction agents. This is especially valuable if the contaminants to be
removed are not soluble or are scarcely soluble in an aqueous ex-
tracting agent and will not disperse in it either.
   The organic solvents which are suitable for this purpose fall  into
two groups: those in which water is soluble  and those  in which
water is scarcely or not soluble. The first category includes acetone,
ethyl acetate, ethanol and isopropyl  alcohol.  Belonging to the
second group are the aliphatic hydrocarbons which  can be consid-
ered as solvents (e.g., hexane). If organic extraction agents are
                                                                                      INTERNATIONAL ACTIVITIES
                                                                                                                             577

-------
                      coflUminiifrf    prttrtitm«fil   soil   •* • •(fiction ~^        ,^ M(nr»tton of         poil-tr«*t  - - "" '"J,V.i   ^ * *[   •'•- _J
                                    T  '
                                    *
                                 Hiving rtlidua
                                     • ilrlding
                                      tgtflt
                                                            Figure 2
                                       Extraction of Contaminated Soil (Detailed Process Scheme)
used, the cleaned soil must  undergo subsequent treatment to en-
sure that  the remaining extracting liquid  is wholly  or virtually
wholly removed.

Cleaning the Extraction Fluid

  A large number of physical, chemical and biological purification
methods are available to clean a contaminated aqueous  extracting
agent. Among them are aerobic and anaerobic biological purifica-
tion, ion exchange, coagulation, flocculation, sedimentation, flota-
tion, membrane filtration, ion exchange, electrodialysis, electroly-
sis, etc. These methods   are extensively used in industry and are
described in detail in the literature. Therefore, the individual purif-
ication processes will not be dealt with here.  For more informa-
tion, refer to the general literature on this subject.
  Which purification process is most suitable in a particular case
depends on many factors: the percentage of clay particles and or-
ganic components in the soil, the nature and concentration of the
contaminants, the composition of the extraction agent and the re-
quirements imposed if the purified extraction agent is discharged
into a sewer system and/or into surface water.
  In general terms however,  the purification  process  is usually
composed of three sub-processes. The first sub-process is aimed at
removing  the colloidal and  suspended particles, e.g., by coagula-
tion, flocculation and sedimentation or flotation. The second sub-
process is aimed at removing  any dissolved organic components,
e.g., by active carbon adsorption. The final step is aimed at de-
mineralizing the extracting agent.
  The sludge resulting  from the purification of an aqueous ex-
traction liquid is usually dewatered mechanically with a centrifuge,
filter press or belt press to minimize the volume of the residue.
  For the purification of a contaminated organic extracting agent,
one  can usually use techniques such as extraction, evaporation,
distillation, stripping, etc. These techniques are extensively used in
the processing industry.  For more information, refer to the rele-
vant literature.

EXTRACTION EQUIPMENT

  Basically, many types  of  equipment are available for the prac-
tical application of the extraction process.  In general, this equip-
ment is already widely  used  in  the  process industry,  in waste-
water purification or in excavation.
  The most critical unit processes in the extraction process are:
•The mixer in which the contaminated soil and the extraction fluid
 are brought into intensive contact with each other
•The separation system for soil particles and extracting agent
•The purification process for the extraction phase
  It will be assumed in the following discussion of suitable equip-
ment that aqueous extraction fluids are employed. This can usually
be expected to be the case in practice.
                                                           Various systems can be used for the intimate mixing of soil and
                                                        extraction fluid.  In  general, these systems involve the generation
                                                        of high shear forces between the particles and between the particles
                                                        and the liquid, the object being to scour off the contaminants pres-
                                                        ent on the surface of the particles and to promote solution or dis-
                                                        persion. These high shearing forces can be generated, for example,
                                                        in a scrubber in which  contaminated soil and a relatively small
                                                        quantity of extraction fluid are  mixed intensively by mechanical
                                                        means. Another possibility for achieving high shearing forces is the
                                                        use of high-pressure liquid jets. When scrubbers and other mixing
                                                        equipment are used, it is often necessary to ensure that the min-
                                                        eral soil particles are not crushed. If many small particles result,
                                                        there can be insurmountable  problems in later stages of the soil
                                                        cleaning process.
                                                           Various systems are available for separating the extraction fluid
                                                        with the impurities in it from  the soil particles. The  main ones are
                                                        discussed below.

                                                        Settlers
                                                           Settlers are usually employed if the settling velocity of the soil
                                                        particles is sufficiently great and differs from the settling velocity
                                                        of any dispersed contaminants present. For the efficient operation
                                                        of the system, several settlers are usually employed in series. Be-
                                                        tween sedimentation steps, the soil  is mixed with relatively clean
                                                        extracting agent.
                                                        Fluidized Beds

                                                           A typical fluidized bed'is shown  in Figure 3. Extraction fluid
                                                        and soil particles are fed in a  counterflow made through a vertical
                                                        column. The superficial  velocity  of  the liquid is regulated so that
                                                        the soil particles settle and can be removed at the bottom of the
                                                        column. The contaminate extraction  fluid (including dispersed con-
                                                        taminants and fine soil particles and dissolved soil components) is
                                                        removed at the top of the column.
                                                        Screw Extractors
                                                           A diagram  of a screw extractor is shown in Figure 4. A screw
                                                        extractor consists of a sloping trough provided with a  transport
                                                        screw. The soil is fed in at the bottom and transported upwards by
                                                        the screw. The extraction fluid is  fed in at the top and flows down-
                                                        wards as the soil particles flow up the extractor.
                                                        Hydrocyclones

                                                           The slurry of soil particles  and extraction fluid are fed tangen-
                                                        tially into the cyclone. The underflow of the cyclone contains the
                                                        soil  particles, still with a small quantity of extraction fluid, while
                                                        the overflow contains the bulk of the extraction fluid with the dis-
                                                        solved and dispersed contaminants in it. For efficient operation of
                                                        the system, a number of hydrocyclones are usually placed in series.
578
INTERNATIONAL ACTIVITIES

-------
                         contaminated soil and
                          extracting agent
contaminated
extracting
agent and
fine particles
      clean
      extracting
      agent






\
(
\




t '
1
t


. \ ' / ' ' ^
t t • t • t ' t ' t 't t
                                            upward current
                                            carrying fine
                                            particles
                                          coarse particles
                                          falling in
                                          downward current
                             clean soil

                           Figure 3
                   Diagram of a Fluidized Bed

Between the separation steps, the  underflow is mixed with (rela-
tively clean) extraction fluid. Hydrocyclones can be. employed for
the separation of particles down to approximately 20 um.
  In addition to being suitable for the separation of soil particles
and contaminated extracting liquid, the above systems can also be
used for post-treatment of the treated soil.
  Aside from these systems, mention should also  be made of a
number of other  separation systems which are possible  but less
feasible in practice, such as:
•Vacuum belt filters, or sieve belt  filters; the contaminated soil is
 put on a conveyor belt, which is sprayed  with extraction fluid;
 the percolated extraction fluid is carried off, cleaned and possibly
 recirculated.
•Rotating liquid sieves; probably  only suitable for cleaning soil
 with particles mainly larger than about 150-250 »im.
•Centrifuges; comparable to hydrocyclones but usually consider-
 ably more expensive.
•Sieve bends; only suitable for separating particles larger than 300-
  Larger particles, still present in the extraction fluid after sepa-
ration of the bulk of the soil, can be removed using hydrocyclones
and/or centrifuges. The process conditions must  be designed so
that only the larger, relatively clean particles are separated leaving
behind any dispersed contaminants which may be present.

FIELD OF APPLICATION
Soil Types

Sandy Soils
  The extraction process is best suited to cleaning soils composed
mainly of sand particles;  it is fairly easy to separate sand par-
ticles from the extraction fluid due to their relatively high settling
velocity. Separation can be achieved using relatively simple sep-
aration equipment, such as settlers. A second reason why the ex-
traction process is highly suitable  for cleaning sandy soils is that
sand particles have a relatively small specific surface area; thus, the
amount of contaminant adsorbed to the sand particles is relatively
low.
  In the case of sandy soils, both aqueous and organic extraction
agents can be used. Which type of liquid is most  suitable is de-
termined  mainly by the  nature of the contaminants.  Adequate
cleaning of the soil usually can be achieved with an aqueous ex-
traction fluid. The presence of a small quantity of clay particles
and/or humus-like substances in the soil imposes no restrictions on
the extraction process. It is, however, to be expected that a large
amount of clay particles and humus will remain in the extract phase
when the soil particles are separated from the extraction fluid and
will end up in the sludge of the purification plant.
  When sandy soils contain small quantities of humus-like com-
ponents, the contaminants are often adsorbed to a large extent  on
these humus materials. Use of an aqueous extraction fluid with a
high pH causes some of the humus-like components to dissolve, re-
sulting  in fairly easy contaminant dispersion in the  extracting
agent.
Loamy and Clay-Like Soils
  In general, loamy and clay-like soils are much more difficult to
clean by extraction than sandy soils for two reasons. First, because
of their small dimensions,  silt  and clay particles  readily form a
relatively stable  suspension with  the extraction liquid.  This  is
especially true for aqueous extracting agents with a high pH. If the
contaminants are present in the extraction liquid as separate small
particles, there  is no  easy way of separating soil particles from
contaminated particles.
  The second reason  that  extractive  cleaning is difficult  is that
many types  of contaminants are readily  adsorbed by  loam and
clay particles. This situation  applies  particularly to  the clay par-
ticles which readily bind ionogenic components, acting more or less
as ion exchangers. Clearly,  contaminants which are bound in this
way to the soil particles will be difficult to remove.
  It is also possible to  extract contaminants from clay and clay-like
soils with an organic solvent. The application of organic extract-
ing agents appears to offer good prolspects for cleaning soils with
organic contaminants  which are insoluble  in water but are soluble
in the organic solvent. However, organic extraction agents general-
ly cost more.
Other Types of Soils
  Generally, what has been said about the  use of the  extraction
process with loamy and clay-like soils also applies to peat and peaty
soils, highly heterogeneous soils and dump sites. In the case of soils
containing large quantities of organic matter (e.g., plant remains
and humus compounds), part of this matter will dissolve or be-
come suspended in the extraction liquid. When the extraction liquid
is cleaned, these components are separated again and finally end
up in the residual sludge.
                                          extracting
                                           agent
                        contaminated
                          eoil
    contaminated
    extracting
    agent and
    fine particles
                            Figure 4
                   Diagram of a Screw Extractor
                                                                                       INTERNATIONAL ACTIVITIES
                                                            579

-------
Contaminants

Heavy Metals and Metalloids

  Two methods are suitable for the extractive removal of heavy
metals (and metalloids) or compounds of these elements. In the
first method, the metallic compounds are dissolved in the extract-
ing agent. In general, aqueous solutions of HC1, H2SO4 and HNOs
are suitable for  this purpose. The action of the extraction agent
can be increased by the addition of sequestering agents. In the case
of amphoteric metals, it is also possible, in principle, to use a  base
as the extraction agent. Bases, e.g., an aqueous solution of sodium
hydroxide, will generally be preferred because of equipment  con-
siderations.
   In the second method,  the contaminants are removed  as solid
particles. A base can be used as the extraction liquid. The function
of the high pH is to  foster the dispersion of the contaminated
particles in the  extracting agent. Water  to which surface active
agents have been added can probably also be used for this purpose.
   A favorable aspect when lye is used is that a substantial part of
the natural organic components (such as plant remains and humus-
like compounds) dissolves. When this happens, the contaminants
bound to this organic matter also dissolve or are dispersed.
   In general, organic solvents seem unsuitable, or  in any case less
suitable, than aqueous extracting agents for the extraction of heavy
metal compounds from contaminated soil.

Free Cyanides and Complex-Bound Cyanides

   Free cyanides often  dissolve well in lye and can, therefore, be
removed by extraction using such a solution. The  complex-bound
cyanides  are usually  present  as  complex-bound  iron cyanides.
Again, extraction can be achieved using lye which dissolves the iron
cyanides.
   In general, it can be assumed that organic solvents are not  suit-
able for the removal of cyanides from soil.

Other Inorganic Components

   For the removal of contaminants such as  acids, bases, phos-
phates, sulphates and nitrates, aqueous extracting  agents are  suit-
able. Here again, a distinction can be made between contaminants
which are soluble in water and contaminants which are not.
   Organic extracting agents are not usually suitable for the removal
of inorganic contaminants.
                          •ludgt


                           Figure 5
        Simplified Process Scheme of the HBG-Extraction Plant
                                                        Organic Components
                                                          In principle, aqueous extracting agents are also suitable for the
                                                        removal of most aliphatic and aromatic hydrocarbons and related
                                                        compounds such as phenols, alcohols,  chlorinated hydrocarbons
                                                        and PCBs.  As in  the removal of heavy metals from soil, a dis-
                                                        tinction can be made here between contaminants which dissolve in
                                                        the aqueous extracting agent and contaminants which do not. The
                                                        first category consists of substances such as ethanol and acetone
                                                        which  can often be washed out of the soil easily with cold or hot
                                                        water free of additives.
                                                          To remove contaminants of the  second category,  it is often
                                                        necessary to enhance the dispersion properties of the extracting
                                                        agent by adding lye or surface active agents. Organic solvents can
                                                        also be used. Which type  of solvent  is  most suitable from a tech-
                                                        nical and  economic standpoint depends largely on the type of con-
                                                        taminant  to be removed, it is  at any rate important that the con-
                                                        taminant to be removed dissolves well in the solvent.
                                                          In summary, extraction  is applicable to virtually all types of con-
                                                        taminations. Sandy  soils  are the most  suitable types of soil for
                                                        cleaning by extraction.
                                                        THE PRESENT STATE OF THE ART
                                                          The cleaning of excavated soil by extraction is a technique which
                                                        is still developing. To date, extraction has been used only on a lim-
                                                        ited scale  and for a limited number of types of contaminated soil.
                                                        In this section, a brief description of the most important develop-
                                                        ments is given, with the emphasis on processes developed or in use
                                                        in The Netherlands and the United States.
                                                          The Hollandsche Beton Groep NV (HBG) has developed an ex-
                                                        tractive cleaning  plant  for   sandy soils in cooperation  with
                                                        TNO.'- '•  w The plant has been operational since the summer of
                                                        1984 and  has a capacity of 25  tonnes/hr. The design is based pri-
                                                        marily on the experience obtained with a similar, smaller pilot plant
                                                        in 1983. A highly simplified diagram of this pilot plant is shown in
                                                        Figure 5.  The following steps occur (the  numbers correspond to
                                                        those in the figure):
                                                          1. Pretreatment of the  contaminated soil to remove large ob-
                                                        jects such as pieces of wood and stones and to break up clods.
                                                          2. Extraction with lye; the primary object is to scour off the con-
                                                        taminants from the soil particles and to dissolve or disperse them in
                                                        the liquid  phase.
                                                          3. Washing of the soil with clean extraction agent in a fluidized
                                                        bed.
                                                          4. The  fine  sand  panicles in the extracting  agent  leaving the
                                                        fluidized bed are separated in hydrocyclones; if desired, this fine
                                                        sand fraction can be rewashed separately.
                                                          5. Drainage of the cleaned soil through a screen
                                                          6. The  spent extraction agent  containing  the contaminants is
                                                        cleaned in a number of steps. Cleaning usually is carried out by pH
                                                        adjustment,  coagulation,  flocculation,  sludge separation, sludge
                                                        dewatering and a second pH adjustment; pan of the purified water
                                                        can be reused.
                                                          The plant  is primarily designed  for sandy soils containing iron
                                                        cyanides (e.g., gasworks sites). Other types of contamination can
                                                        also be removed from soil under certain conditions. For example,
                                                        sandy  soils contaminated  with arsenic  and chlorinated hydrocar-
                                                        bons were successfully cleaned in this pilot plant.
                                                          Ecotechniek BV has developed a plant for hot water washing.
                                                        According to the company, the plant has a maximum capacity of
                                                        approximately 100 tonnes/hr and  is particularly suitable for sandy
                                                        soils contaminated with oily compounds.1'' For the plant to oper-
                                                        ate properly, these compounds have to be removed from the sand
                                                        particles using hot water.  In the past, the plant has been used for
                                                        cleaning beach sand contaminated with crude oil.
                                                          The firm Mosmans-Heidemij has developed a flotation process.
                                                        It was used on a pilot-plant scale  in 1983 on sandy soils contam-
                                                        inated with oily materials (including PCAs).1-I0 A larger plant is
                                                        expected to become operational this year.
                                                          The flotation process is shown diagrammatically in Figure 6. The
                                                        heart of the  plant is a series of flotation cells in which the soil is
580
INTERNATIONAL ACTIVITIES

-------
                   contaminated
                       soil
                                     sieving residue
   flotation
    agents
                                                   foam
                                                (contaminated
                                                    residue)
                     clean soil
                           Figure 6
         Simplified Process Scheme of the Flotation Process
                      (Mosmans-Heidemij)

 cleaned in two or three steps. The contaminated soil is mixed with
 water and a small quantity of surface active flotation agents. It is
 then fed into the flotation cells as a slurry. The function of the flo-
 tation agents is to increase the hydrophobic properties of the con-
 taminants. In the flotation cells, air bubbles continually rise, adher-
 ing to the slurry contaminants  and transporting them to the sur-
 face. By skimming off the layer of foam on the surface, separation
 is effected between the soil and the contaminants.
  Although the process is fairly simple in terms of equipment, it
 is complicated  because the cleaning process is sensitive to varia-
 tions in soil type and types of  contaminants. Each contaminated
 site should be investigated beforehand to establish the optimum
 process conditions (type of flotation agents, concentration of flota-
 tion agents, pH, process times, etc.). In principle, the process seems
 to be suitable  for  cleaning  sandy  soils contaminated  with  oil
 products, organic-chloride compounds, heavy metals and cyanides.
  Bodem Sanering BV has developed a plant to remove contam-
 inants from soil  with high-pressure waterjets.1' " In this plant,
 which has a capacity of approximately 15 tonnes/hr, soil is trans-
 ported through a curtain of waterjets with a water velocity of about
 220 m/sec.  These waterjets detach the contaminants from the soil
 particles. The method is particularly suited to sandy soils contam-
 inated with substances which are insoluble  in water (e.g., oil).  A
 number of successful trials have been carried out.
  In cooperation with the Hollandsche Beton Groep NV, TNO has
 carried out an investigation into the possibility of cleaning sandy
 soils contaminated with organic bromine compounds such as tri-
 bromoethene, tetrabromoethane and higher bromoalkanes.1'3> 8| 20
 The investigation, carried out on  laboratory,  semi-technical and
 pilot-plant  scales, resulted in  a method of  extractive  cleaning
 through which the organic bromine content of the soil was reduced
 from over 100 mg/kg to less than 1 mg/kg. The design of the clean-
 ing plant is  shown in Figure 7. The method of operation generally
 corresponds to  the HBG pilot plant mentioned earlier. However,
 screw  extractors  are used  instead of fluidized   beds,  and the
method of cleaning the extraction liquid is slightly different.
  The above review of extraction cleaning techniques is not com-
plete.1' 20 It does not include all  the Dutch companies and research
institutes which are engaged or wish to engage in extractive clean-
ing of excavated soil. Moreover,  there are presumably develop-
ments about which no information has yet been made public.

U.S. Technology

  A number of developments in the extractive cleaning of soil are
also taking place in the United States. Two plants will be mentioned
here.
  The USEPA has developed a mobile extractive cleaning plant.12
The prototype has a capacity of approximately 5 tonnes of soil an
hour. The soil is cleaned in four steps in a counterflow made with
water as the extraction agent. At each step, the  slurry of soil and
liquid is intimately mixed in a  simple tank by rising  air bubbles.
The water is purified by activated carbon adsorption,  with supple-
mentary purification steps if necessary, and then recirculated.  Re-
garding applications, the plant is said to be usable for many types
of contaminants.
  A hot water  "fluidization" process for cleaning  oil-contam-
inated beach sand was built and tested by the University of Cali-
fornia,  Santa Barbara, around 1970.13 The process is a variation
of the hot water method used in  the Athabasca Tar Sands De-
posits and utilizes water at 95 °C in a ratio of 1.2 m3 to 1 tonne of
soil in a fluidized, upflow bed contactor. The oil is removed from
the fluidized bed with the water overflow and separated from the
water in an oil-water separator. The water is recirculated.
  Tests performed with a sand mixture containing 1 to  2% of a 23 °
API crude oil showed that more than 95% of the crude oil could be
removed. Operation with a 14 ° API residual oil was less satisfac-
tory. The limitations of the process  are connected primarily with
the range or distribution of sand particle sizes that can  be fluidized
without excessive elutration.
  In the United Kingdom, Robertson Research International Ltd
has developed a mobile "dense media cyclone plant" for use on
small-scale mining and dump retreatment projects.14 Separation of
contaminants is based on differences in the specific gravity of the
particles. This plant can probably also be used for some old, haz-
ardous  mine waste sites. The plant  has successfully  processed a
lead/zinc/barite/fluorite prospect and a zinc/fluorite deposit. The
capacity is approximately 15 tons/hr.
  In Germany, a process for the extraction of heavy metals  from
dredged materials by an acid treatment has been examined." The
method was  developed  for  the decontamination of harbor sedi-
ments but is  probably also applicable to contaminated soils. This
technique,  referred to as the "leaching method of Muller,"  com-
prises three steps: (1) an acid treatment with HC1 to extract heavy
metals;  (2) separation of the solids from the solvent;  and (3) the
removal of the heavy metals from the extracting agent by a hydrox-
ide and carbonate precipitation.

PROCESSING OF RESIDUES
  When soil is cleaned by extraction there is always a residue con-
taining  concentrated contaminants.  This  residue  often must be
treated as a hazardous waste. If an  aqueous extraction agent* is
used, the residue comes free as  a sludge. In addition to the actual
contaminants, the sludge usually contains a high percentage of clay
particles and organic soil components (such as plant remains and
humus-like compounds). The fraction of such matter in the sludge
is generally many times higher than that of the contaminants. The
amount of sludge produced per tonne of treated soil, therefore,
largely depends  on the  composition of the soil. The amount of
sludge is also affected by the type of purification process used for
the spent extracting agent: whether chemicals have been added and
the way in which water is removed from the sludge are particularly
important.
  Further processing of the sludge, regarded as a hazardous waste,
can occur in two ways.  The first method involves the transpor-
tation of the sludge to a controlled disposal site. Here, the sludge
•This section deals exclusively with residues resulting from cleaning with an aqueous extraction
agent.
                                                                                      INTERNATIONAL ACTIVITIES
                                                           581

-------



^

lupply »tUf
(lOm'/li)
Hi OH

V i"
[ 	 Nt OH
/1$1^r3

                                            toi,- to*
                                           lUctrotyt*
                                                            Figure 7
                  Process Scheme of the Proposed On-Site Treatment Installation for Removal of Organic Bromine Compounds
can undergo chemical or thermal treatment to immobilize the con-
taminants to a greater or lesser extent.  Transportation to a con-
trolled disposal site will be necessary if no (economically  feasible)
processing method is available. The second  possibility for sludge
handling is to further concentrate and/or destroy the contami-
nants. In principle, various techniques are suitable.17' "  Most of
them are also used, or can be used, for the processing of other types
of chemical waste. The most important techniques are:
•Incineration
•Chemical oxidation  with  the aid of hydrogen peroxide, ozone,
 potassium permanganate, sodium hypochlorite and calcium hypo-
 chlorite
•Hydrolysis in an aqueous environment at increased temperature
 and pressure; this process is usually carried out in an acidic (sul-
phuric acid or hypochloric acid) or  an alkaline (caustic soda or
 caustic potash) environment.
  The aim of the first two processes is the complete oxidation of
the contaminants. The last process is designed to decompose the
contaminants chemically to smaller molecules.  The process can be
aided by the addition of oxygen. Treatment in an aqueous  environ-
ment at  elevated temperature and pressure  with  the  addition of
oxygen is known as  wet air oxidation. However, hydrolysis does
not necessarily result in substances less toxic to man and the en-
vironment than the original contaminants.
  The success of the above-mentioned methods of treatment of
different types of sludge is discussed in the next four sub-sections.

Sludges Containing Heavy Metals (and Metalloids)

  Basically, destruction of  heavy metals is impossible. A  substan-
tial reduction of the volume of the residual sludge can  be achieved
by incineration. In  the  incineration  of a heavy  metal-containing
sludge, attention must be paid to the evaporation of heavy metals.
A number  of heavy metals (or metalloids) easily evaporate (e.g.,
mercury, arsenic and cadmium) and have to be removed from the
exhaust gases.
  A  general problem in the incineration  of sludges obtained in
liquid extraction of contaminated soil is the  entrainment  of small
particles  by the exhaust gases. These particles can be primarily
comprised of heavy metals  and consequently have to be removed
from the exhaust gases.
Sludges Containing Free Cyanides

  Free cyanides can be destroyed by chemical  oxidation (e.g., by
sodiumhypochlorite). A precondition, however, is that the cyanides
are more or less soluble in an  aqueous liquid. For cyanides,  in-
soluble in cold water, thermal treatment methods such as hydroly-
sis at  higher temperatures and incineration are probably the most
promising methods."
  Incineration should occur under strictly controlled process con-
ditions (excess oxygen, relatively long residence time in the after-
burner, high temperature to avoid the formation of toxic com-
ponents. Adequate safety control measures  should be taken to
avoid calamities.  As far as  known,  there is  little large scale ex-
perience with these methods.

Sludges Containing Complex Cyanides (Iron Cyanides)

  From the  literature, it appears that both hydrolysis and inciner-
ation are applicable  when  the  right  process  conditions  are
chosen."• "•  "  The incineration process is possible using a high
temperature, excess oxygen and a relatively long residence time in
the afterburner. Hydrolysis  is possible at temperatures of 250°C
and higher and a residence time of several hours.  However, prac-
tical experience on a large scale is not yet available.

Sludges Containing Organic Compounds

  Incineration currently appears to be the most appropriate treat-
ment method. This is especially  true when different types of or-
ganic contaminants are present in the sludge. Incineration is an es-
pecially attractive method for  treating sludges containing large
amounts of oily compounds.
  The incineration conditions depend mainly  on the types of con-
taminants. In general, halogenated hydrocarbons (e.g., PCBs and
pesticides) require more stringent conditions  of incineration than
simple aliphatic compounds (e.g., oil).
  Numerous organic  hazardous wastes are currently  being des-
troyed by incineration on a large scale world-wide. The knowledge
and experiences gained in this field will be very valuable for the in-
cineration of residual  sludges from  the extraction of soils con-
taminated with organic compounds.
  It is also known that a large number of organic hazardous waste
compounds can also be destroyed by wet air oxidation." However,
at this moment there is a lack of  practical experience on a large
scale.
  Finally, it  has to be noted that  pre-treatment of the sludge by
neutralization,  drying, oxidation,  milling or  stripping of volatile
compounds  can sometimes  increase the technical and economic
applicability of the above-mentioned treatment methods.

COSTS

  The costs  of cleaning a given quantity of contaminated soil by
extraction depend greatly on the type of soil, the nature of the
contamination  and the desired degree of cleaning. The principal
cost factors include:
•Cost of the extracting agent* (chiefly the cost of chemicals to be
 added
*ln this section only aqueous extracting agents are considered.
582
          INTERNATIONAL ACTIVITIES

-------
•Cost of cleaning the extracting agent
•Cost of further processing and/or disposal of the residual sludge
•Energy costs
•Costs of transport, construction and dismantling the purification
 plant
•Interest and depreciation costs of the plant
•Labor costs
•Costs of analyses for process control
•Costs of any additional measures (in connection with safety, pre-
 vention of groundwater and  air pollution and prevention  of a
 noise nuisance)
  Of the above factors, interest on, and depreciation of the invest-
ment are by far the most important.  These costs depend greatly on
the number of hours per annum that  the plant is in operation.
Other important cost factors are labor costs and the cost of pro-
cessing and/or disposal of the residues.
  On the basis of the experience obtained with pilot plant investi-
gations and actual cleanup operations, the costs of extractive clean-
ing are estimated at $20 to $75 (U.S.)/tonne of soil treated. In view
of the lack of sufficient practical data, this cost estimate should be
regarded as approximate.
 CONCLUSIONS
   Extraction can be considered to be a technique which is capable
 of removing all types of contaminants from soil, provided the con-
 taminants can be sufficiently transferred to the extracting agent.
 Thus far, the applicability of the technique has been investigated
 only for the cleaning  of sandy soils with  an aqueous  extraction
 fluid. Favorable results have been obtained for a number of con-
 taminants, including complex (iron) cyanides, aromatic hydrocar-
 bons, oily compounds and organic bromine compounds.
   The quantity of residual sludge formed in the extraction process
 can be a problem. This sludge, in which the contaminants are con-
 centrated, must generally be regarded as a hazardous waste. Addi-
 tional costs are usually associated with the processing of this waste
 in order to destroy the contaminants or substantially reduce the
 quantity of waste.
   The principal techniques available for processing residual sludge
 are incineration, chemical oxidation  and hydrolysis. Practical ex-
 perience on a large scale has been obtained only with incineration.
   Very roughly, the costs of extracting contaminants from sandy
 soils are estimated at $20 to $75 U.S./tonne of soil.
   Of the various techniques for cleaning contaminated soil, the
 ones which are most developed and have  the widest application
 are extraction and thermal treatment. Although the areas of appli-
 cation of the two techniques partially overlap, extraction is prob-
 ably the only cleaning technique suitable for soils contaminated
 with heavy metals.
   To further technical and economic optimization of the  extrac-
 tion process and widen the area of usefulness, research is needed
 in the following areas:
 •Minimization of the quantity of residual sludge formed
 •Processing of residues from the extraction process. For example:
 thermal treatment of sludge containing iron cyanide;  hydrolysis
 of sludge containing iron cyanide;  thermal or hydrolytic treat-
 ment of sludge containing halogenated hydrocarbons (e.g., PCBs
 and pesticides).
 •Extraction of loamy and clay-like soils
 •Extraction of peat  and peaty soils
•Development of processes in which organic extraction agents can
 be used. Attention will have to be paid not only to the  choice of
 organic extraction  agents but also to the treatment of  the spent
 extraction agents.

REFERENCES

 1.  Handbook bodemsaneringstechnieken (in Dutch), (Handbook of Soil
    Treatment Techniques), Ministry of Housing, Physical Planning and
    Environment,  Directorate-General for Environmental  Protection,
    Staatsuitgeverij, The Hague, The Netherlands, 1983.
 2.  Proc. National Conference on Management of Uncontrolled Haz-
    ardous Waste Sites, Washington, D.C., Oct., 1981.
 3.  Proc.  National Conference on Management of Uncontrolled Haz-
    ardous Waste Sites, Washington, D.C., HMCRI, Nov., 1982.
 4.  Handbook for Remedial Action at Waste Disposal Sites,  Municipal
    Environmental Research Laboratory,  Office of Environmental Engi-
    neering and Technology,  Office  of Research  and  Development,
    USEPA, Cincinnati, OH, 1982.
 5.  Kitchens,  T. and Smith, J., "Nations unite to study waste-site cleanup
    routes," Chem. Eng., May 16, 1983, 20.
 6.  Edwards,  R.E., Speed, N.A. and Verwoert, D.E., "Cleanup of chem-
    ically contaminated sites," Chem. Eng.,Feb. 21,1983, 73-81.
 7.  Beoordeling en aanpak van bodemverontreinigingsgevallen (in Dutch),
    (Assessment of and measures against cases of soil  contamination),
    Ministry of Housing, Physical Planning and Environment, Director-
    ate-General for  Environmental  Protection,  Leidschendam,  The
    Netherlands, 1982.
 8.  Hollandsche Beton Groep NV (HWZ-Bodemsanering), Vanadiumweg
    5, Amersfoort, The Netherlands.
 9.  Ecotechniek, Beneluxlaan 9, Utrecht, The Netherlands.
10.  Mosmans-Jeidemij,  Zuiderparkweg  284,   's  Hertogenbosch,  The
    Netherlands.
11.  Bodemsanering Nederland  BV, Daelderweg 15, Nuth, The Nether-
    lands.
12.  Ludwigson, J., Proc. National Conference on the Control of Haz-
    ardous Materials Spills, Milwaukee, USA, 1982.
13.  Mikolaj, P.G.  and Curran, E.J., A Hot Water Fluidization Process
    for Cleaning Oil-Contaminated Beach and Oil Spill Cleanup, Sym-
    posium, USA, 1972.
14.  Robertson Research International Limited,  'Ty'n-y-Coed', Llanrhos,
    Llandudno, Gwynedd, LL30 1 SA, N. Wales, UK.
15.  Muller, G., "Chemical decontamination: A  concept for the final
    disposal of dredged materials and sludges contaminated by heavy
    metals." Heavy metals in the Environment,  2.  International Con-
    ference, Heidelberg, Germany, Sept., 1983.
16.  De grote  schoonmaak  (in Dutch), (The  big  cleanup),  Report on
    symposium of Arp., 1083. Practical studies, Study Association on the
    Department of Civil Engineering of the TH-Delft, The Netherlands,
    Sept., 1983.
17.  De Renzo, D.J., Unit Operations for Treatment of Hazardous Indus-
    trial Wastes, Noyes Data Corporation, Park Ridge, NJ, 1978.
18.  Lehman, J.P., Hazardous Waste Disposal,  Phenum Press, New York,
    NY, 1983.
19.  McBride,  J.L. and Heimbuch,  J.A., "Skid Mounted System Gives
    California Hazardous Wastes a Hot Time,'' Pollut. Eng., July, 1982.
20.  Ontwikkeling Bodemreinigingstechnieken (in  Dutch),  (Development
    of Treatment Processes for Contaminated Soil), P. Symposium Proc.,
    Apr.  1983, Ede, Ministry of Housing,  Physical Planning and Environ-
    ment, Directorate-General for Environmental Protection, The Nether-
    lands, April, 1984.
                                                                                       INTERNATIONAL ACTIVITIES
                                                            583

-------
  MEASUREMENT  OF LOW PERMEABILITY COEFFICIENTS
               BY  MEANS OF  ELECTRONIC INSTRUMENTS

                                 WILHELM GEORG COLDEWEY, Ph.D.
                                     Westfalische Berggewerkschaftskasse
                                           Westphalian Mining  Fund
                                   Bochum, Federal  Republic  of Germany
INTRODUCTION
  For determining the permeability coefficient (kf) of non-cohesive
soils, there is a great variety of methods which have been used in
the laboratory for decades. However,  the study  of the water-
permeability of cohesive soils  with low permeability coefficients
poses problems.
  Low permeability coefficients are an important criterion for con-
structing waste sites. The Institute for Applied Geology of the
Westphalian Mining Fund is often involved in investigating poten-
tial disposal sites and assessing them.  This process  requires deter-
mination of the permeability coefficients of very slightly permeable
soils.
  In order to study permeability coefficients of slightly permeable
soils(kf =  10-7m/stokf = 10 -» m/s), the Institute for Applied
Geology had intended to acquire suitable laboratory apparatus. In-
tensive  efforts to  acquire   a  measuring device from reputable
laboratory equipment makers both at home and abroad were un-
successful. They therefore decided to construct a measuring device.

REQUIREMENTS FOR THE
MEASURING DEVICE
  The following requirements were established:
•Measurement of small permeability coefficients
•Automatic recording of the measurements
•Simulation of the specific overburden pressure
•Simulation of high pore water pressures
•Simulation of high pore water pressure differences
DESCRIPTION OF THE MEASURING DEVICE
  The measuring device consists of  a consolidation cell, a con-
solidation press,  a water tank, recording equipment, a computer
and two displacement transducers (Fig. 1). The soil sample is placed
in the pressure cell and is loaded by the press. The consolidation of
the sample is recorded with the help of a transducer.
  Water is forced up through the sample from below, under known
pressure from  a cylindrical tank with variable volume (piston). The
volume of water flowing through the sample is determined by the
free cross-section of the tank and the displacement of the piston.
The displacement of the piston is measured by means of an elec-
trical displacement transducer. The pressure on the soil sample and
the water pressure can be continuously controlled by the recording
equipment and measured by pressure transducers. Additionally,
the water  temperature at the sample is measured (Fig. 2).
  In order to  measure the permeability of contaminated water as
well, the consolidation cell and the water tank were constructed of
stainless steel.
  All the measurements of the recorder are fed into the computer
at regular intervals by means of a time switch. The calculator is pro-
grammed to  calculate the permeability coefficient automatically
from these measurements.
                                              / tr—19
                          Legend
 I. Frame (lillible through 90' tnd 180*)
 1. Sample
 3. Consolidation cell
 4 Plunger plale
 5 Pressure cylinder
 6. Compressed air line lo pressure cylinder
 7. Consolidation press frame
 H Hller plaies
 9 Cylindrical Unit
 10. Pore pressure gauge
 11. Temperature gauge
 12. Displacement transducer for measuring the consolidation or the sample
 13. Displacement transducer for measuring the volume of water flowing through the sample
 14. Compressed air line lo cylindrical lank
 IS. Recording equipment
 16. Computer
 17. Indicator for the pore pressure gauge (10). temperature gauge (11). displacement transducer on
   cylindrical tank (13). switchable
 18. Indicator for the pressure of the consolidation press
 19. Compressed air line (inlet)
 20. Paper strip (results)
                         Figure 1
   Diagrammatic Sketch of Measuring Device for Water-permeability.
   (designed by Birk, Coldewey, Oeiersbach for the Westphalian Mining Fund.)
584
         INTERNATIONAL ACTIVITIES

-------
                         Figure 2
     Photograph of the Water Permeability Measuring Device
ADVANTAGES AND DISADVANTAGES
OF THE METHOD
  The measurement of the permeability coefficient by means of
electronic instruments offers numerous advantages.
  With this measuring device, a large number of samples can be
quickly processed. The equipment is completely self-sufficient and
can be used over extended periods.
  Furthermore,  the method  permits the simulation of an over-
burden pressure—in a continuous manner—to a depth of 200 m.
This option is especially important since the permeability coeffi-
cient varies with depth.
  A further advantage of the equipment is that one may control
very closely over a wide range of pore pressures and thereby control
the hydraulic gradient  at the cell-inlet. The range of available
pressures is between 5 and 80,000 cm of water at a sample length of
2.5 cm. Using this capability,  one can ascertain at  which pore
pressure the permeability coefficient tends towards zero, where the
transition point from the pre-linear zone to a linear one occurs, or
at what point a  turbulent  flow (post-linear  range)  or similar
behavior arises. In fact, it would also be possible to ascertain where
the sample approaches hydraulic soil failure and at what point it is
eventually reached.
  Finally, one should note that the flow through the sample can be
from the top, from the bottom or horizontal. This change in direc-
tion  can be achieved by rotating the consolidation cell in a frame.
This capability is important because the direction of flow affects
the permeability coefficient.
  A  disadvantage of utilizing electronic instruments is that logging
the results—as compared with other processes—is removed  from
the direct control of the operation.

CHECKING THE MEASURING DEVICE

  Intensive experimentation was undertaken before putting the ap-
paratus into use.

Pressure Transducer

  To check the consolidation press, a calibrated testing ring was in-
corporated in the consolidation cell frame. The pressure on the
frame was measured in two ways:
•By  means of a dial gauge  on the testing ring
•By  means of a pressure transducer with electronic indicator  on
 the  recorder of the permeability apparatus
  Comparison of the two series of measurements showed a linear
relation between the values  derived  by the different two  pro-
cedures.
  Since the compressive stress on the sample surface as well as the
absolute force, is of interest, the compressive stress (a) in N/mm2
was  also calculated (Fig. 3).  Under natural  conditions this stress
depends on the thickness (h) and density (p) of the overburden.
         a = p • g • h                                      (1)
  To simulate the overburden pressure at the consolidation cell, the
thickness of the overburden with densities p = 2 or 2.2 g/cm3 was
calculated and is shown on the lower abscissa (Fig. 3). The available
consolidation press can simulate overburden pressures up to 200 m.
                                                         COMPRESSIVE STRESS
                                                      INDICATION FROM DYNAMOMETER RING (N)
                                                  1000    .000   ^ooo    »wc    >ooo   tooo   tooc
                                                 *IVO    II M   'I M   91 K   104 BO   '« M
                                                    EQUIVALENT THICKNESS OF OVERBURDEN (
                                                           Figure 3
              Checking the Pressure Transducer on the Consolidation Press and Converting for the Specific Overburden Pressure
                                                                                   INTERNATIONAL ACTIVITIES
                                                          585

-------
Pressure Transducer

  The pressure transducer on the water tank was checked with a
calibrated manometer. The measured  values were fully in agree-
ment with those shown on the recorder of the apparatus.

Electronic Displacement Transducers
  The displacement transducer on the water tank was checked with
a dial gauge. The results are shown in Figure 4. The small variations
of the measured values arise from inaccuracies in the mechanical
dial gauge, as was apparent when the dial gauges were compared.
  Then the displacement transducer on the press was compared
with  the one on  the water tank. Here, too, there was complete
agreement.

Volumes of Water
  The volume  of water that  flows through  the soil sample  is
calculated from the free cross-sectional area of the water tank and
the electronically  measured displacement of the piston. The water
emerging from the sample was collected and weighed. Comparison
of the results of these two processes showed complete agreemenl
(Fig. 5).

Computer Program
   The data were  processed with the aid of a Hewlett Packard HP
97 programmable desk computer. The functions of the computer
can be checked with a test program. This checking of the computer
is performed at regular intervals.
   Checking the program on magnetic cards is also necessary, since
the  programs,  too, can be disturbed by external causes (e.g.,
magnetic fields).
 PROBLEMS AND APPLICATIONS
   The equipment  can be readily operated with the help of an
 operating manual. Errors can arise only when, at the beginning, the
 pore pressure and then the consolidation pressure are applied. The
 sample can be pressed out of the cell.
   The range of application lies between approximately kf = 1 .
 10-' m/s and kf =  l« 10- " m/s. Results may be obtained within
 an hour; in fact, even quicker at values such as kf = 1  • 10~6 to kf
 = 1 •  10-8 m/s. From kf = 1  • 10-'°tokf =  1 • 10-'*m/s, ap-
 proximately 1  day is required  to get useful results. Permeability
 coefficients  smaller than kf  =  1 • 10 -I2 m/s can  only be deter-
 mined by rough approximation.
   Reproduceability between kf =  1 •  10~6 m/s and kf =  1 • 10-'
 m/s is high up to the decimal point. This was proven with isotropic
 samples. However, the accuracy drops as values approach kf = 1 •
 10" "  m/s, where it becomes only half of a power of  ten. Beyond
 that kf value, one  can only estimate results within  an order of
 magnitude.
   Inaccuracies  or  errors arise through evaporation  of the pore
 water. This loss of water causes superficial drying-out of the sam-
 ple. As a result, more water gets transported through suction than
 through the actual  flow. This problem becomes more  serious near
 the boundaries, since the process is always accompanied by a tem-
 porary shrinkage of the sample in this zone.
   To avoid this phenomenon, the complete cell is placed in a water-
 bath, thereby eliminating the suction pressure. This step will help to
 avoid the above mentioned inaccuracies.
   Dye tests have shown that boundary flow can be ignored. Even in
 non-cohesive materials, the flow velocity, in the immediate vicinity
 of the  boundary, is only 1/3 higher than in the rest of the sample.
 Thus, calculated over the surface, this phenomena can be ignored.
                    DISPLACEMENT TRANSDUCER
                   /
                         n-lh MEASUREMENT
                                                                                   • VO«.U«t CAl'.AAIEO 9» *f 
-------
TEST RESULTS
  One may best interpret the results  by comparing results with
measurements made using different equipment. Samples were ex-
changed with the Niedersachsisches Landesamt fur Bodenforshung
NLfB (Lower Saxony Geological Survey).
  A permeability coefficient of kf = 1 • 10-7m/s was obtained for
pure quartz dust using the measuring device of the Westphalian
Mining Fund  with a consolidation pressure of 0.5 kg/cm2 and a
pore pressure higher than 0.1 bar;  the value obtained by the NLfB
was kf = 1.6 • 10 ~7 m/s. For other quartz dust samples at higher
overburden pressures,  the agreement was as good. A greater dif-
ference was obtained when natural samples were compared. It ap-
pears that the error  occurred  because a more jointed sample was
put  into the measuring device,  which  then  yielded a higher
permeability coefficient.

EXTENSION OF THE MEASURING DEVICE
  To further improve the accuracy when studying highly cohesive
soils, one may use a narrower water tank. At the  other end of the
scale, a larger water tank is used for permeability coefficients above
kf = 1 •  10-6 m/s.
  Additionally,  new   pressure  transducers  and  displacement
transducers with more precision were installed.  In order to test
samples of different length,  a longer consolidation cell was also
constructed.
CONCLUSIONS

  In conclusion, one can say that the measuring device enables the
following:
•Measurement of permeability coefficients from kf =  10 ~6  m/s
 tokf = 10 ~12 m/s
•Measurement of flow from the top, the bottom and horizontally
•Simulation of overburden pressure
•Simulation of high pore water pressure
•Simulation of high pore water differences

REFERENCES

1. Schmidt, R., "Determining the Permeability Coefficient of Cohesive
  Soils," Report of the Westphalian Mining Fund, 43, Bochum, Aug.,
  1983,  3.
2. Birk,  F., Coldewey, W.G. and Geiersbach, R., "The New Measuring
  Device of the Westphalian Mining Fund for Determining the Water-
  permeability of Cohesive Soils," Report of the Westphalian Mining
  Fund, 43, Bochum, Aug., 1983, 5-9.
3. Benner, L.H. and Coldewey, W.G., "Checking the Measuring Device
  for Determining the Water-permeability of the Westphalian Mining
  Fund," Reports of the Westphalian Mining Fund, 43, Bochum, Aug.,
  1983,  11-18.
4. Benner, L.H., "Physico-chemical Studies of Cohesive Soils with Par-
  ticular Reference to Permeability," Reports of the Westphalian Min-
  ing Fund, 43, Bochum, Aug., 1983, 19-167.
                                                                                    INTERNATIONAL ACTIVITIES
                                                         587

-------
          THE  UPWARD MIGRATION OF  CONTAMINANTS
                          THROUGH  COVERING SYSTEMS

                                              R.M. BELL, Ph.D.
                                           G.D.R. PARRY, Ph.D.
                                        Environmental Advisory Unit
                                              Liverpool University
                                         Liverpool, United  Kingdom
INTRODUCTION

  The majority of contaminated land sites or landfills in the United
Kingdom are isolated or encapsulated. This sometimes requires the
use of several layers of cover materials. The covering systems are
likely to be required to perform three main functions:
•To prevent exposure of the population at risk
•To sustain vegetation
•To fulfill an engineering  role such as accommodating  uneven
 settlement, trafficking, etc.
  The ability of any covering system based  on soil or soil-related
minerals to fulfill the above functions will depend on a number of
factors.2
•Control of upward  and  lateral migration  of contaminants
 through the ground
                      •The ability of the cover material to immobilize pollutants through
                       chemical and physical absorption
                      •Its effectiveness to control water ingress and thus leachate pro-
                       duction
                      •The interaction between covering systems, the contaminants and
                       the biology, e.g., plant root systems
                      •The engineering behavior  of the  system and its component ma-
                       terials
                        In this paper, the authors describe two investigations in which
                      constructed soil columns were used to assess the upward migration
                      of contaminants from contaminated ground into clean imported
                      barriers and covers.
                        In experiment 1,  the columns contained a lead/zinc waste which
                      was covered by various barriers  and topsoil. In experiment 2,
                      various wastes  arising from the coal carbonization process were
                      covered by compacted clay  and topsoil.
                     Topsoil
                   Break Layer
                   Contaminated
                     Waste
                                       •10%
                                  	PVC pipe lined
                                   with black polythene
                   Sand
                                     -PVC pipe


                                     — Water table
                                     —Capillary matting
-Polythene-lined
 trench

- Nylon mesh
                         Figure 1
                  Section of Single Column
LEAD/ZINC WASTE EXPERIMENTS

Materials and Methods

  Soil profiles were constructed in 30 cm diameter PVC columns.
The columns contained 30 cm of topsoil overlying a 30 cm barrier
layer placed on the waste (Fig. 1) to simulate a minimum treatment
for  land reclamation.' The barrier layers  consisted of materials
which are commonly used for this purpose in the United Kingdom:
sea-won sand, land-won sand, clay, pulverized fuel ash (PFA) and
building rubble. The waste, from a  former lead/zinc  mine, con-
tained lead, 32.9  mg/g; cadmium,  0.42S  mg/g; and zinc, 56.6
mg/g, at pH 7.4. Each column treatment and a control, consisting
of a column completely filled  with topsoil, were replicated three
times, resulting in a total of 18 columns.
  The topsoil was sown with a mixture of perennial ryegrass and
white clover,  which is  a common seed  mixture for  amenity
grassland. The columns were maintained in drought conditions and
received only sufficient water to sustain  vegetation growth. All col-
umns were maintained with a constant water table at the base. The
replicates of each treatment were randomized in a polyethylene tun-
nel at the University of Liverpool Botanic Gardens, Ness, in April,
1980, and  monitored  continuously until  dismantled  after  30
months. The establishment of the columns has been described pre-
viously.1
  Soil samples, taken at 5 cm intervals up  the column, were col-
lected at the end of the experimental period and analyzed by atomic
absorption spectrophotometry. The distribution of plant roots at 5
cm intervals was also assessed.
588      INTERNATIONAL ACTIVITIES

-------
                                           Seasand barrier
                                             I16I1)
                                                             200
                                                             MM
                           PFA  barner
                            178:2)
                                                                / / /.
                                        100     1000    BOOO      10      100     1000    COOO  10      100

                                                      Zinc concentrations   mg / Kg
                                                                          Landsand bamer
                                                                            (30s1l
                                           D
                                                                                               Soil
                                                                                               Barner
                                 10      100    1000    BOOO     10      100    1000
                                                       Zinc concentrations mg/Kg


                                                              Figure 2
           Zinc Concentrations (original concentrations) in Barriers Overlying Contaminated Waste, Kept Under Drought for 30 Months.
                                     N.B. It Was Not Possible to Sample the Building Rubble Barrier
                                                          300
                                           Seasand barrier
                                                          600
                           •;    0     10     20    30     1,0
                                                                              300
                                                                     PFA barrier
                                                                              600
                                                             0     10    20
                                                   Proportion of total root weqht in profile (%)
                                                            0
                                           Clay barrier
                                                           300
                                                           600
                                                                    Landsand barrier
                                     10     20
                                           D
Sal
                                                                                          Topsal control
                                                                                 0     10     20    30     I
                                                                                                 J
                                                                                           Building rubble barner
              0    10    20        0     10     20    30
    Proportion of total root weight in profile (%)

                \7/\ Bamer           HH Waste
                                                             Figure 3
          Plant Rooting Pattern in Topsoil and Various Barriers Overlying Contaminated Waste, Kept Under Water Stress for 30 Months
Results and Discussions
  Changes in the concentrations of lead, zinc and cadmium were
found within the barrier layers after the 30  month experimental
treatment. In all cases, however, the barriers prevented transfer of
metals to the topsoil.  The largest increase in metal concentrations
in the barriers was found in zinc concentrations immediately above
the waste (Fig. 2) in the clay and PFA barriers where zinc had in-
creased from 80 to 5860 mg Zn/kg and from 78 to 6620 mg Zn/kg
respectively.
  The patterns of lead and cadmium in the barrier layers were
similar to that for zinc, but less obvious. Small increases in cad-
mium concentrations  occurred only in the lower 5 cm of the bar-
riers; again,  greatest increases occurred in the clay and PFA bar-
riers. Increased lead concentrations were found in the first  10 cm
above the waste. At  the end  of the  experimental treatment, the
plant root distribution at 5 cm intervals was assessed (Fig. 3).
                       In general, the root distribution in all soil columns was similar
                     and, as expected, decreased with depth from the top of the column
                     until the barrier layer was encountered. Some rooting took place in
                     the barrier but little occurred into the waste. In the clay and PFA
                     barriers,  and to some extent in  the land/sand barrier, increased
                     rooting occurred at the base  of the barrier and immediately above
                     the waste.  Zinc concentrations in the established vegetation were
                     also affected by the 30 month experimental treatment, with the clay
                     treatment showing most uptake (Fig. 4).

                     COAL CARBONIZATION  WASTE
                     EXPERIMENT

                     Materials and Methods

                       Soil columns similar to  those  described in the first experiment
                     were established in August,  1982,  within a  polyethylene house at
                     the University Botanic Gardens.
                                                                                       INTERNATIONAL ACTIVITIES
                                                                                 589

-------
    550
    500
    200
g>
I
>•  150
TJ
 X
  CJ>
     100
      50
         SEASAND  LANDSAND CLAY
                                       PFA. BUILDING
                                           RUBBLE
CONTROL
1000

r>o
450


ttttt
•>-•
• —
tv-
•--
"•..


SOI
Cur

WASTE

Tolutnt
Eitracl %
026
10 23)
0 19
0 21

<•• 3

                                                                                            Cycloheiane Water Sol
                                                                                            Łilrjcl%   Sulphate*
                                                                                           .001
                                                                                           10 1)
                                                                                           0 01
                                                                                           --
•—
•--•

	 060
SOIL (0 231
f 	 0 36
• 001 62 <2 «'
(0 1) 1561 <<2) HI
003 70 «2 «'
	 0 20 «0 01 79 <2 «'
i---- 0 15 <0 01 65 <2 <<
	 123 37 67 250 '200
**STC ( ) Qncjmal topsoil concentrations
Concentrations in mcj/Kcj unless stated
                                                                                               Figure 8
                                                                        The Spatial Relationship of the More Significant Contaminants
                                                                             with the Soil Column Containing Clinker and Tar
                                                                   the  soil matrix and  thus would be likely to migrate. This was,
                                                                   however, not shown (Figs. 5 to 8). Some cyanides at low concentra-
                                                                   tions are biodegradable, and it is possible that this has masked any
                                                                   increase in concentration up the column.
                                                                     Only one  experimental column contained significant quantities
                                                                   of phenols (Fig. 8),  and under the conditions of this experiment
                                                                   they did not appear to migrate. The rooting densities at various
590
        INTERNATIONAL ACTIVITIES

-------
                           Table 1
      The Original Analytical Details of the Wastes Used in the
            Second Experiment Involving Tarry Wastes
Sulphate Clinker &
waste spent oxide

pH%
Acid Sol. Sulphate %
Water Sol. Sulphate %
Elemental S To
Sulphide
Total CN
FreeCN
Phenols
Ammonia
Toluene extract %
Cyclohexane extract %
Coal tar
Pb
Zn
Cd
0
1.4
6.1
3.88
11.6
4
6333
17
3
860
13.8
1.7
<500
180
47
0.1
0
2.4
3.3
2.3
32.3
4
467
...
<2
2700
35.8
1.7
<500
529
181
0.7
Tarry Spent
waste oxide
0
5.4
0.36
0.07
0.11
11
53
...
727
443
7.0
2.6
5.7%
1823
350
1.3
0
2.0
5.3
2.71
48.3
5
3%
—
<2
5430
60.6
2.0
<500
63
50
0.3
Note: All Concentrations in mg/kg except those specified in %.

depths down the soil columns are shown in Figure 9. As expected,
the vast majority of roots were found in the topsoil, but the rooting
density decreased  with depth down the column.  Some roots did
penetrate all wastes;  in particular, the clinker and spent oxide
waste.

DISCUSSION

  Under the experimental conditions utilized, contaminants have
been shown to move into constructed barrier layers from underly-
ing contaminated materials. In the first experiment, significant con-
centrations of zinc moved 10 cm upwards into both clay and PFA
barriers over the 30 month experimental period. In the second ex-
periment, acid and water soluble sulphates moved 40 cm through a
clay barrier over the  12 month experimental period. Other con-
taminants, such  as lead, cadmium, cyanides, phenols and coal tars,
did  not move as far.  Such upward migration has also been iden-
tified in attempts to reclaim metalliferous tailings. In British Col-
umbia, a gravel layer was needed to stop  upward migration of
acids, soluble salts and iron aluminum zinc and copper into im-
ported overburden.3
  Upward migration has occurred because of various related fac-
tors: the texture, structure and compaction of the barrier layers has
allowed capillary rise,  and the evapotranspiration of the established
vegetation has created a net moisture deficit in the upper soil levels.
The optimum barrier material within a covering system intended to
prevent upward  migration of contaminants should therefore have a
large particle size so that the large pore sizes will tend to break any
capillary rise.
  The materials or soils above this barrier should be of sufficient
depth and water holding capacity to be able to fulfill any needs of
the established vegetation under designed environmental conditions
(for example, the 1 in 50 year drought). In many cases, it therefore
appears that clays are ineffectual barrier systems, and the establish-
ment of vegetation above the clay would further undermine  its effi-
ciency as a barrier.
  In all  soil columns, the roots of the established grass/legume
sward reached the underlying contaminated wastes. In the second
experiment, the roots had reached a depth of 60 cm in 12 months.
In the first experiment, the roots not only reached the wastes, but
the established vegetation also contained elevated concentrations of
the supposedly isolated metals, particularly zinc, a known phyto-
toxin.
  In many cases, the plant roots increased in mass per unit volume
either just above or just within the waste as compared to the adja-
cent materials.  Contaminant uptake  through  root sorption is
therefore likely to be greater than expected assuming the more nor-
mal gradual decrease in rooting density with depth.
  There  are many questions that need to be answered concerning
the relationship between the conditions within  the  columns and
within the polyethylene tunnel house as compared to conditions oc-
curring  at  a reclaimed site.  It  is likely  that the upward fluxes
achieved in the soil columns were so high that they might never oc-
cur under the climatic conditions prevalent in the United Kingdom.
There is also the possibility that rainwater would wash the migrated
contaminants back down the soil profile. This effect would depend
on the cation exchange capacity of the barrier or topsoil. Cation ex-
change derives from the negative  charges  on clay micelles and
organic matter which provide a binding capacity for metals and
other positive ions. The stability of these resulting complexes is,
however, questionable.
                          Sulphate Clinker & Clinker Spent Oxide
                                 Spent Oxide & Tar
yU->3

60*65

30-35

r
• —
• 	
*"'
•--•

	 \i-i
SOIL

	 2 (,
CLAY

	 0-005

LI- 1

3 k

30



1 7

0 1



3 7
n ?

1 3

                            Figure 9
         Root Densities within the Soil Columns (mg/g dry soil)
  The columns provided a means of measuring the upward water
and   contaminant  flux  under the  extreme  conditions  in  a
polyethylene tunnel house. Where vegetation growth  was kept
high, and the water table kept at the base of the column, upward
fluxes would be extremely  high. While information relating these
conditions to those of reclaimed site needs to be collected, the con-
structed soil column remains a useful technique in assessing poten-
tial covering systems and the long term efficiency of remedial ac-
tions.

REFERENCES

1. Jones, A.K., Bell, R.M., Barker, L.J. and Bradshaw, A.D., "Cover-
   ings for Metal Contaminated Land," Proc. National Con/,  on Man-
   agement of Uncontrolled Hazardous Waste Sites,  Washington, DC,
   1982.
2. Lutton, R.J., "Evaluation of Cover Systems for Solid and Hazardous
   Wastes," NTIS PB81-166340, 1982.
3. Ames, S., "Reclamation of Land Disturbed in Mining, Proc. 3rd An-
   nual British Columbia Mine Reclamation  Symposium,  Vernon, BC,
   1979,311-32.
                                                                                     INTERNATIONAL ACTIVITIES
                                                          591

-------
    SYNOPSIS  OF 1983-1988 OUTLOOK  OF ENVIRONMENTAL
     CONCERNS  FROM  SCIENTISTS AROUND THE WORLD*
                                               WILLIAM J. LACY
                                                    Consultant
                                               Alexandria, Virginia
                                             ROBERT F.  HOLMES
                                    U.S. Environmental Protection Agency
                                                Washington, D.C.
 INTRODUCTION

  This paper contains a summary of the remarks made by each
 participant in a panel discussion which the senior author chaired on
 September 23,  1983, at the 4th International Conference for En-
 vironmental Participation at Paul Sabatier U., Towlance,  France.
  The comments made by each speaker were personal in nature and
 not officially those of his or her particular country, university or
 government agency.
  The following is a short synopsis of each panel member's re-
 marks.

 Dr. EIGohari, Egypt, National Research Center, Cairo, Egypt
  Dr. Gohari   was concerned that there are definite limitations in
 the design and operation of wastewater treatment plants  regarding
 the destruction of toxic compounds and viruses. Research and
 development should be focused toward:
 •Recycling waste water to reduce overall water withdrawal
 •Use of less harmful chemicals
 •Increased use of physical/chemical treatment instead of biological
 treatment.

 Dr. Robert Martin, University of Birmingham,  Birmingham, Eng-
 land
  Dr. Martin stated that since the middle 1970s, the world has wit-
 nessed an energy crisis, industrial recession and increasing sophis-
 tication of analytical instrumentation. Meanwhile, standards are
 being formulated based on measurements made to  ever higher
 orders of magnitude. With these standards, pressure groups exert
 untold influence and induce hysteria and fear in  the uninformed.
 A proper perspective must be maintained so that instruments are
 controlled, and are not the controlling factor, so that decisions re-
 garding environmental standards are logical, economical  and tech-
 nically feasible.

 Dr. B.A. Bollo Csito (Australia)
  Dr. Bolto stated that there was a need for new water and waste-
 water treatment processes. In developed countries, lower capital
 costs would characterize these new processes; in developing coun-
 tries,  the emphasis would be on appropriate technology. Greater
 attention should be given to the recycling of products  from waste-
 waters and to the removal  of heavy  metals from  wastewater
 sludges.  The  significance of water supply   for agriculture in
 Australia meant that methods of removal of  salts and turbidity
 should be explored.
Dr. A. Hamza (Egypt), Institute of Public Health
  Dr. Hamza pointed out that the monitoring of wastes was a
major concern worldwide and that, in the future, there should be
more attention paid to the monitoring of industrial and domestic
wastewater treatment processes.  Industry from developed coun-
tries was welcome in Egypt, but overseas aid, employment and
commercial growth should not obscure  responsibility for the im-
pact of hazardous wastes on the environment.
Dr. R. Ben Aim (France), University Paul Sabatier
  Dr. Ben Aim's opinion was that wastewater treatment com-
monly involved the treatment of dilute aqueous systems. Research
should consider the behavior of micropoUutants in dilute systems.
The increasing use of physical/chemical processes in wastewater
treatment and the increasing use of biological processes in water
treatment have resulted in an increasing similarity in both water
and wastewater treatment. Greater emphasis on the optimization of
treatment plant operation should be a priority.

Dr. L. Pawlowski (Poland), Loblin Marie Curie S Kowdowski U.
  Dr. Pawlowski stated  that treatment  processes should become
cheaper and more efficient.  To meet these objectives, it would be
necessary to acquire a greater understanding of the mechanisms by
which these  treatment processes removed particular pollutants.
This understanding could only result from greater chemical knowl-
edge of the processes, the pollutants and their inter-relationships.

Dr. G.N. Pandey (India), National Environmental Engineering
Research Institute
  Dr. Pandey told the group that  numerous research objectives
were  necessary  in  India. The significance  of agriculture  was
stressed. Research should investigate the environmental impact of
pesticides and fertilizers;  a  balance between energy use and en-
vironmental degradation should be sought in the production of
food. Industrial waste treatment, corrosion research, the develop-
ment of low energy technology and the need  to improve commun-
ication  between  environmental scientists and the medical pro-
fession to better understand the effects of organic pollutants on
human systems were all important.
Dr. L.H. Wong (Taiwan), Taichung U.
  Dr. Wang highlighted areas of concern in Taiwan. The extreme-
ly high population density means that waste disposal is of major
importance; disposal of solid wastes is hindered by a shortage of
 •Specifically from: Egypt, United Kingdom, Australia, France, Poland, India, Taiwan, Italy, Israel, Soviet Union, Thailand, Japan, Cuba, Chile, Canada, Korea and Belgium.

592      INTERNATIONAL ACTIVITIES

-------
land. Air pollution problems have been encountered; the presence
of highly toxic  dioxin has been  observed. Nuclear  energy plants
have discharged nuclear wastes into seawater; research should con-
sider the polluting  effects of such wastes on seawater and  its
ecology.
Dr. L. Liberti (Italy), Bari
  Dr. Liberti reported on two-on-going research projects in Italy.
One project involves the study of the interception of wastes from
towns on the Adriatic coast; ammoniacal nitrogen and phosphates
present in effluents are being converted to ammonium phosphate
fertilizers. The potential benefits of this project are two-fold: fertil-
izer production and cleanup  of  the Adriatic  Sea.  In the second
project, scientists are studying the reuse  of wastewater for  agri-
culture; large ponds are used to  collect and store effluents.  Sun-
light has been observed to inactivate viruses.

Dr. U. Zoller (Israel)
  Dr. Zoller felt that the essential issue was not what could be done
in future years; it was what should be done now. Decisions on what
should be done depend on economy and various social constraints.
The future  should see greater emphasis on  environmental  educa-
tion and ultimate disposal of pollutants;  inadequate information
is presently available on the final effects of pollutants on the world
ecosystem.
Dr. V. Soldatov (USSR), Minsk
  USSR  Academician  Soldatov welcomed the conference and
noted that it had become a tradition for  environmental scientists
throughout the world, since the conference's beginning in the mid-
1970s, to focus on  only one aspect of technology (i.e., ion ex-
change) and then examine all the ramifications to the solution of
environmental problems.
  Dr. Zimny Poland stated that the emission of oxides of nitro-
gen and sulphur had significantly increased in the last 30 years. Re-
search was necessary to investigate the effects of such air pollution
on soil and the plants and animals living in and on that soil.
Mr. L. Roland (UK), Foster-Wheeler, Reading, Berks, G.B.
  Mr.  Roland reminded the assembled delegates that he, as a rep-
resentative of industry, was in a minority being surrounded by aca-
demic  researchers. The economic facts of life in industry  meant
that  future research and development should be conducted with
economic realism  and without  courting hysteria from  pressure
groups and the media.

Additional Comments

  Additional comments by other delegates present at the discussion
session  included remarks by Prof. A.L. Kowal (Poland), who
wanted to see more research on the utilization of wastewater; re-
use of water and recovery of products would be an investment for
the future. Dr. A. Trier (Chile) pointed out the need for air  pollu-
tion standards in developing countries. Various unidentified speak-
ers from other countries including Thailand, Japan,  Cuba and
Canada streesed the need for the use of non-toxic products in agri-
cultural applications, the need for research on methods to clean up
marine and estuarine environment, the significance of acid precip-
itation arising from air pollution, the  importance  of international
cooperation  in the fight  against environmental pollution  and the
need for effective low cost  solutions to air, water and solid  waste
problems.
CONCLUSIONS
  The authors feel that developing nations, not having hard cur-
rency, need environmental pollution control techniques based on
indigenous resources. Their environmental goals should be indexed
to their development vis-a-vis the industrialized nations.  For ex-
ample, if   one plots the  U.S. environmental protection criteria
against its industrialization or economic development since the turn
of the century,  one finds the increasing level of environmental
standards directly proportional to  the rise in  economic develop-
ment and industrialization.
                                                                                    INTERNATIONAL ACTIVITIES
                                                         593

-------
                  OVERVIEW OF HAZARDOUS  WASTE  SITE
                                     PROBLEMS  IN  WALES

                                               RONALD A.  PAGE
                                       Water and Environmental Division
                                                   Welch Office
                                        Cardiff, Wales, United Kingdom
INTRODUCTION

  Wales is a small country which, with Scotland, Northern Ireland
and England, forms the United Kingdom. It has a population of
just under 3 million people and a land area of 8,018 square miles.
Although politically united to England for nearly 450 years with the
two countries sharing common systems of law and government,
Wales has a history, culture and language of  its own. Although
English predominates, Welsh, a Celtic language, is thriving.
  The Welsh Office which is responsible for  functions in many
fields such as agriculture, education and environmental control has
substantial  administrative autonomy. Within  the Welsh  Office
headquarters in Cardiff, the Water and Environmental Protection
Division advises on a range of matters including hazardous waste
sites. This Division has a research program which is designed to
identify, alleviate and possibly provide solutions to environmental
problems in Wales.


DEFINITION OF CONTAMINATION

  "Contaminated land" is the term used in the United Kingdom to
describe land that contains toxic substances in such concentrations
that they present a potential threat directly or indirectly to man, to
the environment or to such other targets as building structures. A
distinction is made between contaminated land and derelict land,
which has been defined as land which is so damaged by industrial
development that it is incapable of beneficial  use without treat-
ment. Examples of the  latter category are disused quarries  and
former railway land; land dereliction from natural causes is  not
covered by the definition. Much, but not all, derelict land is con-
taminated; some sites, although contaminated, are in beneficial use
and are therefore not derelict.
  Land can be contaminated in a number of  ways, e.g.,  by at-
mospheric fallout, by flooding or seepage of liquids, or by deposi-
tion or spreading of solid contaminants.  However, the concentra-
tions of toxic  substances  so produced are in most cases not high
enough to constitute an immediate danger; in Wales, the term con-
taminated land is  considered to embrace land which has been con-
taminated by metalliferous mine wastes,  industrial usage or waste
disposal.
  Although not containing toxic substances, a type of derelict land
which presents a potential hazard to man  and other targets is land
upon which colliery waste and  slate  waste have been  dumped in
such a way that the land constitutes a danger to those living nearby.
This type of derelict land has a particular significance for Wales
because of the disaster which occurred in  1966 when colliery waste,
made unstable  by heavy rain, avalanched down the mountain to
                                                     engulf Pantglas School in the village of Aberfan. One hundred and
                                                     forty-four  people, including  116  children,  died  within  a  few
                                                     minutes. This tragedy, more than anything else, dramatically
                                                     underlined  the dangers of spoiling land. Since 1966, a determined
                                                     drive has been underway in  Wales to clear hazardous sites and to
                                                     remove unsightly waste so that land can be restored to productive
                                                     and beneficial use.
                                                     HISTORICAL BACKGROUND

                                                       Metals were mined in Wales before the Roman occupation, and
                                                     the Roman conquest of the Principality was possibly caused by the
                                                     need of the Romans for Welsh lead and gold. Even as early as the
                                                     18th century it was said that lead mines "enrich a person or two in
                                                     an age and entail poverty on hundreds for generations to come. The
                                                     waters from the mines  spread sterility over the adjacent fields and
                                                     kill all the fish in the rivers." Mining reached its peak in Wales in
                                                     the periof from 1845-1938, but has now  virtually ceased except for
                                                     speculative attempts to remove metals from mine wastes on a com-
                                                     mercial basis. However, mines  being worked at the time of the
                                                     Romans are still giving cause for concern today  because they con-
                                                     stitute a possible threat to health and are a cause of poor fish sur-
                                                     vival in some Welsh waters.
                                                       The industrial revolution also left Wales with a  legacy of con-
                                                     taminated land. At a time when people  were struggling to make a
                                                     living, it is understandable that consideration for the environment
                                                     took second place. Although much has been done to make good the
                                                     damage caused by earlier generations, Wales still shows the scars of
                                                     the industrial revolution. A notable example is the Lower Swansea
                                                     Valley which was the most important non-ferrous smelting center in
                                                     the world during the 18th century. Refining of copper,  lead, silver
                                                     and zinc left the Valley with wastes from 250 years of working, but
                                                     reclamation in progress from 1966  has  done  much to  restore the
                                                     area.
                                                       Wales also was used as a dumping ground for toxic chemicals for
                                                     many years before the disposal of wastes was controlled: therefore,
                                                     there are a number of sites, such as former  quarries,  containing
                                                     drums of unknown chemicals which are a permanent threat to the
                                                     surrounding area. Such sites present complex problems making it
                                                     almost impossible to devise a satisfactory method of redevelopment
                                                     within a reasonable cost.
                                                       While despoliation of the land was ignored, and understandably
                                                     so, by past generations, there is a new environmental consciousness
                                                     in Wales today. The words Lower Swansea Valley, an area in South
                                                     Wales, are virtually synonymous with industrial  devastation—such
                                                     places, tolerated in the past, are no longer acceptable.
594
INTERNATIONAL ACTIVITIES

-------
RECLAMATION
  Within a month of the Aberfan disaster, a special unit was set up
in the Welsh Office to lead, encourage and coordinate a program of
reclamation. A determined effort has been made to clear away
dangerous dereliction  and to  restore land  to productive and
beneficial use. Since many land reclamation projects are designed
to create land for new industry, the functions of reclamation were
transferred to a separate body, the Welsh Development Agency,
which came into operation on Jan. 1, 1976. The Agency has power
to meet the whole cost of reclamation schemes promoted by Coun-
ty and District Councils in cooperation with other public bodies
such as the  National Coal Board, Forestry Commission and the
Countryside Commission.
   In its first three  years of operation, the Agency approved pro-
jects for the reclamation of 6,000 acres of land at a cost of more
than   43 ($55 US) million. To come more up to date, the Welsh
Development Agency announced  a Ł35 ($45 US) million rolling
program in August 1984 as a determined attack on the worst prob-
lems remaining in  Wales. The program,  to be completed pro-
gressively over the next  5 years, gives top priority to schemes
necessary to eliminate  hazard to  life  and  property. Reclamation
releasing land for economic development receives the next priority
and the program includes many such schemes.
RESEARCH

   With large areas of Wales bearing the scars of centuries of min-
ing, quarrying, smelting and other  past industrial activities,
mounds of coal and slate waste, toxic heaps of mine spoil, quarries
containing highly toxic chemicals  and the decaying remains of in-
dustrial processes can have a marked effect upon the lives of people
living nearby.  The Welsh Office has a research budget with funds
available for study of these problems to clearly identify their extent
and nature and to develop methods of reducing if not removing the
hazards. A brief description of some of the research which has been
conducted or is being conducted  now on the problems  posed by
hazardous waste sites follows.
Halkyn Mountain Project
   Mines in the Halkyn Mountain area of North East Wales were
worked at the time of the  Romans. Their operation  left con-
taminants, and this  research  project arose because of the public
concern about possible health hazards presented by the high level
of metals in  local soils. Starting in 1975, a study was carried out to
ascertain the relationship  between blood lead levels of the  local
population  and the lead content of soil, dust,  locally grown
vegetables, air and water.
   As expected, the concentrations of heavy metals in the soils were
found to be high, particularly when compared with concentrations
in a collection of soils from another part of Wales with no known
contamination (Table  1). Settled house  dusts and garden  soil
samples  were  taken at houses where blood samples were taken:
metal concentrations were found to be high in the soil and the dust
(Table 2).
   Although no simple figures can be given  for acceptable concen-
trations of metals in soil and dust, it is interesting to note that it is
generally accepted that only concentrations for lead in soil below
500 /tg/g can be considered as uncontaminated for the development
of the land  for garden purposes in the UK. Some think the same
figure  to be the safety limit for lead in dust.
   In view of the public interest in the UK in lead, the study concen-
trated  on this metal. Therefore,  blood samples were taken from
young children and adult women  to determine the extent to which
lead from the soil was getting into the bodies of residents in the area
(Table 3).
   The blood lead levels were found to be 30 to 50% higher than for
women and  children in other parts of Wales.
   Although  these  levels  were high enough to  constitute an im-
mediate threat  to the health of the local residents, it  was felt that
the study should try to  discover how lead was getting into the
bodies of the residents.
                           Table 1
     Summary Values for Total Metals in 260 Soil Samples from
   Halkyn Mountain with Values from West Wales for Comparison
                    Metal Concentration
                       ( ug/g dry aoil)
Halkyn Pb
Mean 886
Mininun 35
Maximum 47,995
Zn Cu OJ
728 18 6.1
10 2.3 0.4
49,393 252 540
 West Wales
 Maxitrun
                     70
                                 195
                                           29
                                                        2.4
                           Table 2
    Summary Data for 59 Paired Samples of Garden Soils (S) and
                Dust (D) from the Halkyn Area

                        Metal Concentration
                           ( ug/g dry matter )
              Pb
 Mean      1779    480   1143   1166


 Median    1474    346    827   937


 Minimum     33      8.4   46   122


 Maximum   9631   2943   4383   5239
i Cu Cd
D
1166
937
122
5239
S DSD
24 200 62 0.8
19 159 4.2 0.7
5.4 15 0.1 0.1
91 893 27 3.6
                           Table 3
     Mean Blood Levels (in /ig/100 ml) in Halkyn and a Control
  Village.  (A Correction Is Made for Capillary Results in Children.)


Halkyn
Adult Women
Mothers
Children
Control
Village
Mothers
Children
Observed Corrected
Mean Mean
Blood Blood
Level Level
12.8
11.8
22.6 14.3

7.9
17.6 11.2
Proportion Over
20 ;ig/100ml 30pg/100ml

5.5%
4.5%
15.9%

0.0%
5.5%

0.0%
0.0%
0.3%

0.0%
0.0%
  As a result, lead in air and the water supply was determined; the
concentrations were so low that they were thought to make only
minor contributions to body intake. An assessment was also made
of the levels in the dust on childrens' hands, on kitchen surfaces us-
ing the wet wipe technique and in locally grown vegetables. A cor-
relation was found between the amount of lead on childrens' hands
and their body lead; there was also evidence that lead could be get-
ting into the food chain through dust on kitchen surfaces.  The
                                                                                     INTERNATIONAL ACTIVITIES
                                                          595

-------
eating of locally grown vegetables was also associated with raised
blood lead levels. Lettuce, carrot, beetroot and radish were found
to absorb more lead than the other species examined, whereas let-
tuce, carrot, swede and cabbage took up more cadmium. Potatoes,
peas, beans and cauliflowers were found to be poor absorbers of
both metals.
  Advice  was therefore given to residents to avoid high intake of
locally grown vegetables such as lettuce and carrots, to keep work-
ing surfaces in the home extra clean, to encourage children not to
suck their fingers and articles such as toys and to wash home grown
produce carefully before consumption.

Background  Heavy Metal Survey for Wales

  To say  that  land and  other  environmental  media are  con-
taminated, it is necessary to show that the concentrations of con-
taminants contained in them are higher than normal. It is difficult
to define what is meant by normal, and little data are available to
provide a baseline for  comparison.  The lack of data  became quite
apparent during the Halkyn Mountain study.
  Consequently, the University College of Aberystwyth was asked
to  determine the  concentration  of heavy  metals  in Welsh  en-
vironmental media so that a baseline could be established for future
comparison.  The program of work started in 1983  involves the
analysis of soils, dusts, grass, cereals and vegetables for arsenic,
cadmium,  copper,  mercury, lead  and zinc.  The results of the
analysis of the first of the 1,640 soil samples taken for the survey
are shown in Table 4.  It is intended that maps showing the heavy
metal distribution in the soil of Wales will eventually be produced
from this work.
Survey of Contaminated Land

  Although it may  be  thought that the locations of contaminated
areas in a small country like Wales would be  well known, this is
only true of the more obvious sites since no comprehensive surveys
have ever been conducted in the UK. Derelict  land surveys under-
taken by local authorities and coordinated by  central government
have only provided estimates of the amount of contaminated  land
in Wales. Large  areas of central Wales are sparsely populated and
yet, in contrast,  the south is congested.
  Land is in short  supply, particularly in the  coal mining valleys
where only a very small amount of it is suitable for development
because it  is too steep. This shortage  of land brings pressure to
develop any unused sites, and land near or in towns is at a premium
in Wales. Failure to discover before redevelopment  that a site is
contaminated can be costly both in terms of financial resources and
in relation to the risks to which the developers and the eventual oc-
cupants of the site  are exposed. Therefore,  a  need was felt  for a
survey which would locate contaminated sites and which would
classify them according to  their potential hazard, the need  for
remedial treatment and the  factors which were  likely  to inhibit
redevelopment.
                           Table 4
  Summary Statistics for Metals Concentration in Welsh Soil Samples
               (Metal Concentrations pg/g dry soil)
                                                                                    Table 5
                                                                      Structure of Site Contamination Record*
                 Pb
                      Zn
                            Metal Concentration
                                Cu       Cd
                                                   CD
                                                        Ni
 Minimum

 Maximum

 Mean

 Median

 N
3.4
3369
93
39
225
276.8
5.5
2119
81
62
225
148.5
<0.5
215
19
13
225
26.3
<0.3
15
0.6
0.3
225
1.18
<0.2
190
9.3
8.2
225
13.3
<0.9
169
19
14
225
18.2
                                                         Field No
                                                                       Name
                                                                                   CMS name
                                                                                                  Description
                                                            1

                                                            2

                                                            3

                                                            4

                                                            5

                                                            6


                                                            7

                                                            8

                                                            9

                                                           10

                                                           11

                                                           12

                                                           13


                                                           14

                                                           15

                                                           16

                                                           17

                                                           18

                                                           19

                                                           20


                                                           21

                                                           22

                                                           23

                                                           24

                                                           25

                                                           26
Code number

Grid reference

Grid sequence

Type

Name

Location


Topography
Contaminants

Period of use

Site area

Contaminant Scale


           toxicity

Proximity to housing

Status
Hazard  factor

Development factor


Confidence

Comments
Last update

Source
COD

OR

GSQ

TYP

NAM

LOC


TO1

TO2

TO3

CON

TIM

APE

CSC


CTX

pro

STl

ST2

ST3

HF

OF


CNF

OD1

002

003

UPD

SOU
 Unique identifier

 htap  reference

 1km2 location

 Site use

 Sit* name or owner

 County, district 4

  nearest town

 Description of site structure
Moot likely contaminants
Period of contaminating use

Site area in hectares
fcnount of contaminated

 material

Toxicity of contaminants
Distance to housing (km)
Current condition of site
Scale of hazard

Scale of priority for

 attention

Probability of contamination

Mdltlonal information
Date of latest check on data
Primary source for site

 identification
                                                          A survey of Wales was  conducted in  1983/84 by  Liverpool
                                                        University and funded equally by the Welsh Office and the Welsh
                                                        Development Agency. A methodology was developed using various
                                                        information  sources  such  as  maps,  aerial photographs, local
                                                        knowledge,  etc.,  to  locate sites which  were  then classified as
                                                        described above.
                                                          The survey, the first in the United Kingdom, took one man-year
                                                        to complete and identified over 700 sites covering a total area of
                                                        3,787 hectares of land which are believed to be contaminated. No
                                                        soil surveys were attempted because of the large number of sites in-
                                                        volved,  but it is apparent that such surveys are needed to fully
                                                        assess the potential problem in  the event of development.
                                                          A deliberate decision was taken not to record sites which were in
                                                        beneficial use, housing for example, because it  was  felt that to in-
                                                        clude them in the survey might create needless concern.
                                                          In addition, sites with  areas less than 0.5 ha were usually ex-
                                                        cluded. To do otherwise would  have led to a street by street survey
                                                        which would have  been  bogged down with an investigation of
                                                        possible sites which had little or no development potential.  Never-
                                                        theless, where acutely  contaminated sites such as gas works and tar
                                                        lagoons smaller than  0.5  ha were identified, they were recorded
                                                        because of their significance.
                                                          Wales now has a comprehensive record of sites over the whole of
                                                        the country which are thought likely to be contaminated. The infor-
                                                        mation about each site is divided into 26 fields (Table 5) and has
                                                        been entered in a computer database so that site  records  can be
                                                        picked out  by means of the reference number or name of the site.
                                                        More importantly, the computer has been programed to give, on
 596
INTERNATIONAL ACTIVITIES

-------
request, all records of sites with a particular contaminant, in a par-
ticular area, with pre-specified hazard or development factors or
combinations of both.  A typical computer record for one site is
reproduced (Table 6).
  The records of this survey can be viewed as a national register
which can be used to forewarn potential developers of the possibili-
ty of meeting problems and therefore avoiding unplanned costs, ac-
cidents and health hazards during and after redevelopment. The
question of a national register, which could be consulted as a stan-
dard practice in land conveyancing transactions,  is  under con-
sideration but has not yet been fully discussed within Welsh Office
because the results  of  the survey  have only recently become
available.

Effectiveness of Land Reclamation Schemes

  Although much progress has been made since 1966 to reclaim
contaminated land in Wales, very little information is available to
assess the cost-effectiveness and life expectancy of reclamation
systems. Liverpool University has been asked by the Welsh Office
to develop models which can be used to assess the effectiveness of
land reclamation systems.
  Initially, an assessment is being made of existing models to deter-
mine which model or models can be developed to predict the op-
timum reclamation system, the life expectancy of such a system, the
cost versus life expectancy options available and the consequences
to the environment of the breakdown of a reclamation system. The
developed model or models will  be used to  predict the optimum
reclamation system for sites in Wales for which monitoring data
are  already available. No results are available as the project only
started in April, 1984.

Mineral Fibres from Slate Waste

  Slate has been worked from the hills in some parts of North
Wales for centuries. The middle of the last century was a boomtime
in Britain for the building industry, and the demand for slate was at
its peak.  In less than 100 years,  thousands of tons of slate were
                          Table 6
   Example of Computer Print-Out of Site Contamination Records



  DMS current record.      Fil« - CONTAMINATED LAND IN WALES....  f-*q

  COD	   12032  OR	  SJ24373V
  390	  SJ2473
  TYP	  CHEMICAL WASTE
  NAM	            BURNINO TIP
  LOC	
  CLWVOl DELVNi  FLINT
  roi	
  LARGE HEAPS ADJACENT TO TIDAL INLET FROM R.DEE
  ro:	
  T03.
  CON	
  SULPHIDES I VOLATILE OROANICS
  TIM	  18»0-l»80           ARE	
  CSC	
  CTX	  HIGH   PRO	    0.4
  9T1	
  DI8U9EDI PARTLY VEGETATED] SOME AREAS BURNING UNDERGROUND
  ST2	
  ST3.
 Hf	   4  OF	   3 CNF.
 COl	
 LARGE AMOUNTS OF WASTE RAYONl PARTLY COMBUSTED,
 COJ	
 COJ.
 UPO	  SEP 19B3
 90U	  L.».
quarried out of the mountainside, split, cut and exported for roof-
ing. In most places today the quarries are silent, but there are many
reminders  of slate's golden age. The huge quarries and the vast
spoil heaps still tower over the slate villages and dwarf all around
them.
   Apart from being an eyesore, this waste constitutes a potential
danger to  people living nearby as its stability must always be in
question. In view of this danger, it was thought that if a use could
be found for the  slate debris, it would encourage its removal. Con-
sequently,  the Welsh Office has funded a project to investigate the
use of slate as a raw material to make fiberglass.
   Fibers have been produced from slate both by drawing to pro-
duce continuous fibers and by blowing to form mineral wools. The
preferred process involves melting the slate and appropriate fluxes
in a platinum crucible at very high temperatures. Commercial ap-
plications are now being considered for the fibers, and interest has
been expressed by industrial firms in their use as a possible replace-
ment for asbestos in some products.
CONCLUSIONS

   The last meeting of the NATO/CCMS Pilot Study  Group on
Contaminated Land was held in Cardiff in April, 1984. The Welsh
Office was very pleased to host this meeting because contaminated
land presents a serious challenge to the Principality. Wales, with its
legacy of hazardous waste sites, provides a microcosm of the prob-
lems  encountered in many parts of industrial Europe. Any in-
itiatives which would improve the chances of redevelopment of
contaminated sites are most welcome to the people of Wales.
   Wales is a green and beautiful land with mountains, lakes and
beaches second to none, but 200 years of industrialization have left
many grim reminders of the past.  In less  than two decades, the
worst of the  accumulated devastation has been tackled  by the
research and positive action which followed the horror of Aberfan.
   However as old problems are solved, new ones arise—recently,
for example, there was concern when the erosion of a crude cover-
ing layer exposed waste asbestos on a former industrial site near a
housing estate. In another case, the  excavation of a site  for in-
dustrial redevelopment revealed gross soil contamination by PCBs
resulting from the disposal  and deterioration of  capacitors. It
seems, therefore,  that much remains to be done before the debt to
the land has been fully repaid.
ACKNOWLEDGEMENTS
   This paper  was written and  is published by permission of the
Welsh Office. Any views expressed in it are those of the author and
do not  necessarily concur with those of the Welsh Office. The
author would like to thank Professor Glyn Phillips, The North East
Wales Institute,  Dr.  Graham  Parry, Liverpool University, Dr.
Brian Davies, University College of Wales, Aberystwyth and Dr.
Peter Elwood of the Medical Research Council whose research for
the Welsh Office  forms the basis of this paper.

REFERENCES
1.  Halkyn Mountain Project Report, published by Welsh  Office,  main
   contributors Dr. Brian Davies, University College of  Wales, Aberys-
   wyth and Dr. Peter Elwood, Medical Research Council.
2.  Liverpool University, Survey of Contaminated Land in Wales, Report
   to Welsh Office, Environmental Advisory Unit.
3.  Davies, B., Background Metal Values in the Welsh Environment, Sum-
   mary of early results, report to Welsh Office.
4.  Slate  Fibre Research Project,  Report to Welsh Office, North East
   Wales Institute.
                                                                                     INTERNATIONAL ACTIVITIES
                                                          597

-------
           TREATMENT AND DISPOSAL OF HAZARDOUS
         WASTEWATER FROM  FOUNDRY  CYANIDE  HEAT
      TREATMENT OF DELTA  STEEL  COMPANY  LIMITED

                                            EDWIN OHONBA
                                         SYLVESTER OBASEKI
                            Water Supply and Environmental Control  Sector
                                      Delta Steel Company Limited
                                           Bendel State,  Nigeria
INTRODUCTION

  In the early 1950s and 1960s, the greatest pollution problem in
the steel industry was the amount of fumes and dust emitted to the
atmosphere.' Today, the emphasis is switching to the problems
caused by solid and hazardous waste and the control of discharges
of liquid effluents.
  The Delta Steel Company began to use cyanides in a new heat
treatment process in the foundry department where the steel had to
be quenched with oil and rinsed with water.
  Cyanides, which are extremely toxic, especially at low pH, result
from the scrubbing of steel plant gases, from metal cleaning and
electroplating processes and  from  certain chemical industries.
Small doses (2.0 to 4.7 mg CN/day) are normally not lethal to
adult humans as the liver is capable of detoxifying the chemical.
However, death will occur  as the result of large doses when the
detoxification processes of the liver are overwhelmed.
  Many lower animals and fish seem to be able to convert cyanide
to cyanate  which  does not inhibit  respiratory enzyme activity.
However, not all fish are immune. Studies have shown that concen-
trations of 1 mg/1 will kill certain pollution-sensitive fish in 20 min,
and compounds formed by the reaction of cyanide with heavy
metals may be even more toxic than cyanide alone. It is for this
reason that the control of  cyanides in industrial  effluents is ex-
tremely important.

CYANIDE  DESTRUCTION

  This process was entirely  new to the investigators since the steel
plant was still in its infancy. Hence, in their preliminary investiga-
tion, the authors had several constraints. In spite of this, the in-
vestigators have perfected a method for the treatment and disposal
of the  plant's cyanide-bearing  wastewater.  In  this paper, the
authors present the results of their preliminary investigation of the
process employed by the Delta Steel Company to remove cyanide
from their contaminated rinse waters.
  The alkaline chlorine oxidation method, which is the most widely
applied process for the destruction of cyanide, was used.
  Oxidation of cyanide occurs by two separate chemical reactions.
In the  first stage, cyanide is oxidized  to carbon dioxide and
nitrogen; this  reactionis pH  dependent. In the  second stage,
cyanide is oxidized to cyanate (CNO~); at pH 9.0, the reaction will
go into completion in 3 min. The reactions are as follows:
  C12 + 2OH-  i=f  OC1- +  Cl- + HOH               (1)
  CN- + OC1- ±=7  CNO-  +  Cl-                     (2)
  CNO- + 1.5 OC1- + 0.5 HOH  —  0.5N2
                  + HCO3~ + 1.5C1                 (3)
                                                   The samples for the laboratory tests were prepared by dissolving
                                                 a  known amount of reagent in water to correspond to values in
                                                 Table 1. Various dilutions were made and treated with calcium
                                                 hypochlorite (65% purity) and caustic soda. The authors used 2kg
                                                 of calcium hypochlorite/lcg cyanide which is about 1.2 times the
                                                 stoichiometric amount required to convert the total cyanide into
                                                 nitrogen and carbon dioxide.
                                                   After alkaline chlorination  treatment according to pH levels
                                                 shown in Table 1, potassium iodide starch paper was used for the
                                                 detection of available chlorine. Then, the resulting precipitate was
                                                 removed via  filtration. The amount  of residual  cyanide in the
                                                 filtrate was then determined by the pyridine-pyrozolone method as
                                                 outlined by the USEPA in Methods for the Chemical Examination
                                                 of Water and Waste-water.'

                                                 Experimental Procedure

                                                   Titration using silver nitrate is used to measure concentrations of
                                                 cyanide exceeding 1 mg/1. This analytical procedure uses a standard
                                                 solution of silver nitrate to titrate cyanide in the presence of a silver
                                                 sensitive indicator:
                                                 •Pipette 100 ml of sample into a 300 ml Erlenmeyer flask
                                                 •Add 6 ml of 10% NH4OH solution and 0.2 g KI
                                                 •Titrate with 0.1 N Ag NO3 solution until turbidity appears as the
                                                 end-point
                                                                                    Inlt1*1 •••pl« *,
                                                                                    (untreated)
                                                      0                 pH v«lu*>                  14


                                                                        Figure 1
                                                   Cyanide Concentration Reduction as Progressively Larger Amounts
                                                              of Calcium Hypochlorite Are Added
598
INTERNATIONAL ACTIVITIES

-------
                           Table 1
                      Cyanide Treatment
Sample Type
Rinse Water
  Cyanide/oil
  extracted mixture
First Treatment:
a. Rinse water
b. Cyanide/oil mixture
Second Treatment:
a. Rinse water
b. Cyanide/oil mixture
Final Analysis:
a. Rinse water
Concentration
    (mgl)
    91.05
    15.61
     7.8
    2.01
     1.0
 PH
 7.8
 5.7
12.50
12.20
12.00
  The presence of turbidity requires correction with a known KCN
solution to determine the blank value; to do this, prepare a SO mg/1
KCN solution, take 10 ml of titrant, calculate the blank correction
and subtract the calculated blank value from the  known sample
value.

Treatment
  For  effective treatment, eight parts of chlorine should be used
with one part of cyanide (Table 1).

CONCLUSIONS
  The  reaction progressed rapidly in the treatment phase. From the
graphical representation (Fig. 1), one can see the sharp decrease in
the  cyanide  concentration  after  the  first  2 kg  of  calcium
hypochlorite were added.
  The  rapid decrease in the cyanide concentration,  as shown in
Figure 1, gives a true picture of the treatment effectiveness.


REFERENCES
1. Manual on Methods for Chemical Analysis of Water and Waste-water,
  USEPA, Washington, DC.
2. Thomas,  M., "Current attitudes towards the control of Pollution in
  Steel Industry," June 1983.
                                                                                      INTERNATIONAL ACTIVITIES
                                                                                            599

-------
     PURIFICATION AND RECYCLING  OF  GROUNDWATER
         CONTAMINATED  WITH PETROLEUM PRODUCTS
 AND  CYANIDES—THE KARLSRUHE  (FEDERAL REPUBLIC
   OF GERMANY)  DRINKING  WATER  TREATMENT PLANT
                                         RIP G. RICE, Ph.D.
                                            Rip G. Rice, Inc.
                                           Ashton, Maryland
 INTRODUCTION

  The Durlacher Wald water treatment plant in Karlsruhe, Fed-
 eral Republic of Germany, is located next to a major railroad mar-
 shalling yard through which considerable quantities of petroleum
 and petrochemical products have been transported for years. Dur-
 ing transfer of these products, many spills have been experienced,
 and these have resulted in pervasive contamination of some of the
 Durlacher Wald plant groundwater wells.
  On the other side of the plant is an abandoned chemical dump
 which, although closed down in the early 1900s, had begun leach-
 ing cyanides into two of the Durlacher Wald groundwater wells.
  By the late 1970s, two of the plant's four water supply wells had
 been closed down because of chemical contamination, and the
 other two wells were being threatened. However, instead of clos-
 ing down this plant, moving to a new site and building a new treat-
 ment plant, the city opted to install a new and unique ground-
 water treatment process, the technical basis for which  had been
 demonstrated  in laboratory studies but which had not been proven
 at full-scale operational water treatment plants.
  The key to the new process involves the use of ozone to partially
 oxidize the biorefractory organic pollutants, rendering them biode-
 gradable. Contaminated water is ozonized, then reinjected into the
 groundwater stream where it is purified by the natural biochem-
 ical action of groundwater microorganisms.  The process, which
 was installed in 1980, has rejuvenated the contaminated ground-
 water supply,  allowing the  Durlacher Wald  plant  to  reopen its
 closed wells and to continue producing high quality drinking water.
  A  detailed report of this success story was published by Nagel,
 el a/.1 in a German language journal. Because of the  unique nature
 of this treatment  process, its potential applicability to a broad
 range of similarly contaminated groundwaters and the fact that the
 process has not been described in the English literature, the present
 author will review this publication, adding a discussion of the tech-
 nical principles which are the basis for the treatment process.

 THE GROUNDWATER CONTAMINATION
 PROBLEM AT KARLSRUHE

  The oldest of several water treatment plants supplying the city
 of Karlsruhe, Durlacher Wald is located next  to the  rail marshall-
 ing yard  and has  pumped water  from four wells (each 32.70 m
 deep) since the turn of the century. Average  daily water produc-
 tion is 33,600 m'/day (8.8 mgd). Groundwater normally is pumped
 to the surface, oxygenated to raise dissolved oxygen (DO) levels
 and sent to the Karlsruhe distribution system. In recent years, how-
 ever, well tt\,  closest to the  rail yard, had become  contaminated
 with  petroleum products and was closed down in 1978. The second
                                               closest well to the rail yard (well #2) also was becoming contam-
                                               inated and was close to being shut down.
                                                At the other end of the well field, the number four well, next to
                                               a woods, had become contaminated with  complexed cyanide from
                                               a chemical plant dump which had been abandoned in 1906; at the
                                               same time, well  #3 also  was showing traces of cyanides. The re-
                                               lationship of the well field to its contaminating sources is shown
                                               in Figure
                                                                                            I
                                                   njjo
                                                      111.00
                                                                     Figure 1
                                                Durlacher Wald Water Works—Relationships of Contaminating Sources
                                                               to the Groundwater Wells
600
INTERNATIONAL ACTIVITIES

-------
  Contamination of the well closest to the rail marshalling yard
was indicated by a sharp increase in organics levels (to > 5 mg/1
DOC), an increase in turbidity, a drop in DO concentration  (to
C,"l mg/1) and an increase in the iron and manganese levels  (to
0.07 and 0.04 mg/1, respectively). The water quality in  well  tfl,
located 150 m downstream of the first well, began to decrease in
the same manner during the first few months after well #1 was shut
down.
  At the same time, well #4 was becoming contaminated  with  the
hexacyanoferrate complex "Berlin blue". Concentrations had risen
above the maximum contaminant level of 0.05 mg/1, sometimes to
as high as 0.10 mg/1. Consequently, well  #4  also had been shut
down.
  Earlier  studies of  the  groundwater hydrology  at this  plant
showed that the wells are fed by groundwaters flowing into  the
wellfield from the direction of the  rail yard and also from  the
woods at the other end of the field.  At the average daily removal
rate per well of 400 m3/hr (33,600 mVday total water removal
from the well field), the flow of groundwater between wells #1 and
#2 is about 0.06 m/day. From these data, the normal flow time
between the two wells is approximately 8 months.
THE CORRECTIVE WATER TREATMENT PROCESS
Technical Basis

  Ozone has been used as  a water treatment oxidant and disinfec-
tant since the early 1900s.  Although it was employed initially as a
disinfectant in the later stages of water processing, ozone today is
used more often to  oxidize contaminants in the early stages of
drinking water treatment.2
  Even though ozone is the strongest oxidizing agent available  for
water treatment, it is rare  that all carbonaceous materials present
can be oxidized completely to CO2 and  water, even  under pro-
longed ozonation conditions. On the  other hand, ozone is quite
capable of partially oxidizing organic materials, cleaving double
bonds to  produce aldehydes, ketones, alcohols and carboxylic
acids. Aromatic rings, can be ruptured by ozone, producing C2-C4
aliphatic oxidation products. High molecular weight organic com-
pounds can be cleaved oxidatively  to produce lower molecular
weight materials.
  All of these oxidized materials contain higher levels of oxygen-
containing moieties and thus have higher degrees of polarity. This
means that many relatively non-polar, biorefractory organic con-
taminants (such as petroleum-based hydrocarbons) can be rendered
biodegradable upon partial oxidation with  ozone.  Several  au-
thors3' 4-5 have confirmed this benefit of  ozonation, i.e., of con-
verting biorefractory organics into biodegradable organics.
  Ozone normally is generated from dried air in concentrations of
1% to 3% by weight; thus when water is ozonized, it is also aerated
quite efficiently. As a result, the dissolved  oxygen  contents of
ozone-treated waters will be increased  simultaneously during ozo-
nation.
  Additionally, ozone readily oxidizes soluble iron and manganese
ions to  higher valent cations (Fe+3 and Mn+4) which hydrolyze
rapidly to produce insoluble materials [Fe(OH)3 and  MnO2,  re-
spectively], which are readily removed  from the ozonized  water by
filtration.
  Finally, ozone will oxidize free cyanide and many complexed  cy-
anides to the less toxic cyanate ion, which further degrades to CO2,
nitrogen and/or nitrate.

Hydraulic Flows
  The new Durlacher Wald treatment  process consists of ozoning
water withdrawn from the contaminated and previously shut down
supply well #1, splitting the volume of ozonized water into five
equal fractions, then reinjecting these fractions into each of five in-
filtration wells. Three of these infiltration wells are placed strate-
gically between supply well #\ and the rail yard; each is about 75 m
from supply well #1, and each is laterally located 50 m from each
neighboring  well. These  three  infiltration wells form a partial
groundwater intrusion boundary zone  between the contaminating
      Infiltration
                            3 V«—intrusion  zone
                  Well
                «o
                  Well  #3
                   o
                  Well
Infiltration
   wells— "
                                         -Intrusion zone
                            Figure 2
  Durlacher Wald Water Works—Positioning of the Five Infiltration Wells
railyard and the supply well field, through which contaminated
groundwater coming from the rail yard now must pass. This barrier
delays penetration  of the contaminated groundwater into the well
watershed and mixes it with ozonized water which contains high
levels of DO and biodegradable organic materials.
  Two other infiltration wells of the same type are positioned 75 m
in front of supply well #4, 50 m from each other, forming a second
intrusion zone between the supply well field and the contaminating
abandoned chemical dump. The positioning of the five infiltra-
tion wells with respect to the four supply wells and their sources of
contamination is shown in Figure 2.

Process Details

  A schematic diagram of the ozonation/reinjection process in-
stalled at the Durlacher Wald water treatment plant is  shown in
Figure 3. Water from contaminated supply well #1 is withdrawn
at the rate of 400 mVhr and ozonized in the water treatment plant
located on the surface at the rate of 1 mg/1 03 per mg/1 of DOC.
This volume of ozonized  water is split into five  equal portions
(80 mVhr each) and pumped simultaneously into the five infiltra-
tion wells described earlier.
  The ozone installation at Durlacher Wald is designed to produce
1.3  kg/hr of ozone from air; however  normal plant operation is
conducted at an ozone production rate of 500 g/hr. If necessary,
ozone  can be produced from commercial oxygen, increasing the
ozone  output to about 2.6 kg/hr. Total energy requirements for
ozone  production,  including air preparation and controls, are be-
tween  18 and 20 Wh/g of ozone produced.  The dissolved oxygen
content of the water after ozonation has been increased to 9 mg/1.
  This ozone treatment process  was installed at  the Durlacher
Wald plant in May of 1980 and began operating in June, 1980.
EFFECTIVENESS OF THE NEW WATER
TREATMENT PROCESS

  During the initial period of operation, June 1980 through August
1980, 450 mVhr of water was pumped out of supply well #1, ozo-
nized and distributed equally (150 mVhr) to the three infiltration
wells ahead of supply well #1. The rate of groundwater flow under
these  conditions is approximately 6 m/day, which means  that
                                                                                    INTERNATIONAL ACTIVITIES      601

-------
                     ConUmtnated Water fr

                        Donor U 8 mg/1) in mid-July.
  During October,  1980, the initial process  was modified so  that
only 80 m'/hr of ozonized water was sent to each of the three infil-
tration wells near supply well #1 (240 m'/hr);  the remaining 200-
240 m'/hr  of the ozonized water was sent to the two infiltration
wells in advance of supply well #4. The decrease in the volume of
water sent  to the three infiltration wells ahead of supply well tt\
resulted in  reducing the groundwater flow velocity, thereby length-
ening the residence time of the infiltrated water in the ground to
approximately 22 days. This resulted in a decrease of DO levels in
supply well tt\ to approximately 3 mg/1. Over the next five months
of operation under  this revised mode, the DO  level in supply  well
#\ again rose. By the end of 1981, the DO levels in all four supply
wells had reached an average level of 6-8 mg/1.
  During the initial  10-week  period of operation of the  new pro-
cess (June-August 1980), oxygen consumption in supply well #1 was
approximately 40 kg/day; subsequently, however, the oxygen con-
sumption became almost negligible. These data indicate that the
initial  high organic loading present  in contaminated supply  well
#\  was being degraded microbially, under aerobic  conditions,
and that the carbonaceous,  polluting  materials were being con-
verted  into CO2 and water. After the organic contaminants  had
been largely  biodegraded, the oxygen  consumption decreased to
values more in line with those of relatively unpolluted waters.
  The  dramatic  changes in DOC levels found  in each of the four
wells prior  to and subsequent to installation of the ozone treatment
process are shown in Figure 4. Levels of 3 to 5.5 mg/I in early May,
1980, in supply wells #1, #2 and #3 fell to about  1.5 mg/1 by August
1980. By the end of 1981, the DOC levels of all four wells were only
slightly above 1 mg/1.
  Similar favorable data were obtained by analysis of specific or-
ganic constituents. Prior to installation  of the  ozonation process,
petroleum  hydrocarbons were found routinely in supply wells #1
and #2; currently, these types of organic compounds no longer can
be detected. Additionally, concentrations of iron and manganese,
which previously had reached levels of 0.07 and 0.04 mg/1, respec-
tively, are again below the limits of detectability.
  Finally, the levels of cyanide  in supply well #4, which  had risen
to as high  as 0.11 mg/1 in early 1980 (and  which was present in
supply well #3 at levels of 0.01-0.02 mg/1), dropped to below the
                                                        German drinking water standard of 0.05 mg/1 by July 1980,  and
                                                        by late 1981 was undetectable in both supply wells #3 and #4.

                                                        Bacteriological Considerations

                                                           Because ozone is a powerful disinfecting agent, it might appear
                                                        that ozonation of contaminated groundwater could  destroy micro-
                                                        organisms and thereby  decrease  microbiological  activity in  the
                                                        groundwater.  However,  such is not the case when  this treatment
                                                        process is properly designed and  operated. Although ozone  will
                                                        disinfect microorganisms in the 400 m'/hr of wellwater withdrawn,
                                                        this water is reinjected into the flowing groundwater stream, which
                                                        is replete with microbiological activity. The use of low ozone dos-
                                                        ages (1 g 03/g DOC), coupled with ozone's short half-life in water,
                                                        assures that there will be no  residual ozone present by the time the
                                                        ozonized water is reinjected into the infiltration wells and comes in
                                                        contact with fresh groundwater.
                                                           The ozonized,  disinfected, reinjected water stream  also con-
                                                        tains  high  concentrations of dissolved oxygen in addition to par-
                                                        tially oxidized organic materials which are more polar  and more
                                                        readily biodegradable than before ozonation. As a result of having
                                                        more DO plus a biodegradable food supply,  microbiological  ac-
                                                        tivity in the groundwaters flowing toward the supply  wells from  the
                                                        infiltration wells actually increases as a result of the ozonation step.
                                                           Total cell counts (living and  dead) in all four supply wells  be-
                                                        fore the ozonation process was installed were between 2.5 and 3.9 x
                                                        lO'/ml. One year later, the total cell counts had risen by a factor of
                                                        about 10 (to between 1.5 and 2.0  x 10'  ml).  On the other hand,
                                                        the high colony counts in supply well »\ before the ozonation sys-
                                                        tem was installed (140-280/ml) become lower by a factor of about
                                                        100 one year later.  These currently low colony counts of supply
                                                        well #1 are of the same order of magnitude as those of the other
                                                        three supply wells.
                                                           Numerous bacteriological  tests  have confirmed  that  the pre-
                                                        ozonation  treatment  has  increased the biological activity in  the
                                                        ground significantly.6-1 This  increased bacteriological activity not
                                                        only has improved the quality of the water, but also its bacteriologi-
                                                        cal quality. Since the decrease in bacterial colony counts in supply
                                                        well #1 was achieved (end of  1981), the city of Karlsruhe has been
                                                        able to send waters withdrawn from supply wells tfl, #3 and #4
                                                        directly to  the city distribution system, without oxygenation, with-
                                                        out additional chlorination and with no increase in colony counts
                                                        being observed in the distribution system.
                                                                                    Figure 4
                                                             DOC Values (Sliding Mean Values) for the Operating Wells at the
                                                              Duriacher Wald Water Works, Before and After Installation of
                                                                               Ozonation System.
602
INTERNATIONAL ACTIVITIES

-------
DISCUSSION

  Elimination of the cyanide concentrations from supply wells #3
and #4 may not be a direct result of ozonation, but rather an in-
direct result. It is well-known that hexacyanoferrate complexes are
very stable to ozonation.8 In  fact, when these materials were used
as photographic bleaches, ozonation was employed to reconvert the
spent  hexacyanatoferrous bleach chemicals to the initial ferric
forms, thus allowing recycle and reuse of the spent photographic
bleaches.
  Injection of ozonized water into the two intrusion wells ahead of
supply well #4 creates a diversion in the groundwater flow pattern
so that most of the cyanide-containing contaminated water now by-
passes supply well #4 (and consequently supply well  #3). However,
soil samples taken in the vicinity of these two intrusion wells and
supply well #4 have shown significant increases  in microbiological
activities.6
  Thus a combination  of groundwater  flow  diversion and in-
creased levels  of aerobic microorganism activity produce  the im-
provements in water quality noted in supply wells #4 and #3. From
the point of view of supply wells #1 and #2, diversion of the flow
of contaminated groundwater from the rail marshalling yard has
been shown not to be the operative mechanism.6'7

CONCLUSIONS

  Groundwaters contaminated with petroleum  chemicals  and cy-
anides were removed from the ground and altered in properties
by  ozonation in such a  manner that their reinjection into the
ground resulted in an improvement in biological activity in the soil.
Biorefractory  organic materials were partially oxidized by ozone,
producing more readily biodegradable materials.
  At the  time of publication (1982),  the Durlacher  Wald water
treatment plant had  been rejuvenated  and had operated  success-
fully  for 18 months. More recent information7 confirms that the
process continues to provide high quality drinking water to the cit-
izens of Karlsruhe four years after process installation.
  Because of the relative simplicity  of the process, it  can be
adapted to changes in operating conditions and to cope with new
contamination problems which may arise. However, it is necessary
to have sufficient  knowledge regarding the flow of contaminated
groundwaters both directionally and with respect  to volumes. In
addition, the identity of contaminating  pollutants should be known
as fully as possible to allow determination of the optimum  amount
of ozone which will be required to convert the pollutants into bio-
degradable oxidation products prior to reinjection.
  The actual purification of the contaminated groundwater takes
place in the ground and is  only stimulated and improved by the
added ozone treatment. Thus, the natural, biological ground puri-
fication processes,  which are improved and optimized using the
added step of chemical oxidation, are employed. As such, the pro-
cess should be considered by municipalities  currently  drawing
groundwater from contaminated aquifers.
  It must be recognized, however, that not all polluting organic
materials can be oxidized, even with ozone, at sufficiently reason-
able rates to allow  them to be converted into biodegradable ma-
terials. For  example,  some  of  the volatile  organic  chemicals
(VOCs) which have been proposed for  regulation by the USEPA
(such as chloroform, carbon tetrachloride and tetrachloroethane)
are relatively unaffected by ozonation.  On  the other hand, other
VOCs (such as benzene, xylenes, di- and  trichloroethylene)  are
reactive with ozone, and this unique treatment process might  be
effective in coping  with VOC problems involving these types  of
chemicals.
REFERENCES

1.  Nagel, G.,  Kiihn, W., Werner, P.  and Sontheimer, H., "Purifica-
   tion of  Groundwater by  Infiltration of  Ozone-Treated Water",
   Wasser-Abwasserl23, 1982, 399-407.
2.  Rice, R.G., Robson, C.M., Miller,  G.W. and Hill, A.G., "Uses of
   Ozone in Drinking Water Treatment", J. Am.  Water Works Assoc.
   73,1981,44-57.
3.  Gilbert, E., "Investigations on the Changes of Biological Degradability
   of Single Substances Induced by Ozonation", Ozone Sci. & Engrg. 5,
   1983, 137-149.
4.  Legube, B., Langlais, B.,  Sohm, B. and Dore', M., "Identification
   of  Ozonation  Products From Aromatic  Hydrocarbon  Pollutants:
,   Effect on Chlorination and Biological Filtration", Ozone Sci. & Engrg.
   3, 1981,33-48.
5.  Somiya, I., Yamada, H., Nozawa, E. and Mohri, M., "Studies on the
   Biodegradability and GAC Adsorbability of Micropollutants by Pre-
   ozonation", in Sixth  World Ozone Congress  Proc.  Norwalk, CT;
   Intl. Ozone Assoc., 1983, 108-110.
6.  Werner, P., Univ. of Karlsruhe, Federal Republic of Germany, Private
   Communication, 1982.
7.  Kiihn, W., Univ.  of Karlsruhe, Federal Republic of Germany, Private
   Communication, 1984.
8.  Lorenzo, G.A. and Hendrickson, T.A., "Ozone in the Photoprocess-
   ing Industry", Ozone Sci. & Engrg. 1, 1979,235-248.
                                                                                      INTERNATIONAL ACTIVITIES       603

-------
AN  OVERVIEW OF SOLID WASTE  MANAGEMENT  IN  CHINA
                                                     GU YOUZHI
                                                    ZHU YAOHUA
                                 Research  Institute of Environmental Protection
                                                    Shanghai, China
 INTRODUCTION

   Solid wastes are inevitable results of productive and consumptive
 activities of human beings. Wastes are generated in the processes of
 exploiting natural resources and manufacturing of goods. Those
 goods also will become  wastes after  being  used or consumed.
 Associated with urban population growth and  production develop-
 ment, the quantities of solid wastes have been dramatically increas-
 ing. The annual generation of residential  refuses and night soil by
 200 million urban people in China amounts to approximately 150
 million tons. Industrial wastes have totaled  370 million tons an-
 nually, and it has been estimated thai the solid wastes may reach 1
 billion tons/yr  around the year 2000.
   In the past, China did not take the management of solid wastes
 seriously; dumping was the sole method of disposal. Over the years,
 540 million tons of solid wastes, covering more than 40,000 ha of
 land,  have  been  accumulated.  Consequently,  17,000 ha  of
 farmland have  been contaminated by salts and heavy metals, thus
 destroying soil  texture.
   "The Environmental  Protection  Law (For Trial  Implemen-
 tation)",  issued in  September 1979, stipulates  that  "rational
 use of natural environment,  prevention  and elimination of en-
 vironmental pollution and damage to ecosystems" should be en-
 sured; active prevention and control of noxious substances from
 polluting and damaging the environment are needed. In China, at-
 tention has recently been focused on the control and management,
 present and future, of solid wastes.

 POLLUTION  IMPACT OF SOLID WASTES

   In China,  the utilization of mineral resources is  at a rate of
 50-60%, and utilization of energy resources at  30%. Recently, total
 quantities of solid wastes generated' were over 500 million tons/yr:
 (unit in million tons/yr)
     Waste Type
     Coal Gangue
     Mineral Tailing
     Cinder
     Fly Ash
     Smelting Slag
     Residues from Chemical Industry
     Residential Refuses and Night Soil
                                           Amount
                                           100
                                           100
                                            70
                                            40
                                            40
                                            16
                                           146
   Dumping requires a vast expanse of land. Owing to management
 imperfections, residues or ashes from some industries were directly
 discarded into water bodies. Since the 1950s, that has caused a 1.3
 million ha reduction in water surface area of Chinese rivers and
 lakes.
   Only 2% of residential refuses and night soil are treated, and un-
 treated night soil contains a large number of pathogenic  bacteria
 that can threaten human health. Therefore, the municipal wastes
 are unacceptable to peasants, and disposal of such wastes is becom-
 ing more difficult.
   Ordinarily, the contamination extends several times beyond the
 area  of uncontrolled  dumping. Not only the soil, but also the
 groundwater will eventually be damaged. It is estimated that there
 was a  9 billion yuan  RMB  (Chinese  People's  Currency) ($3.9
 million, US) annual loss as a result of environmental pollution and
 the failure to recover useable materials.
   An example of this  is  a 1,800  ton  heap of residues  near a
 Shanghai Zinc Smeltery. The heap contains 0.03-0.2% cadmium
and the dump site covers 0.7 ha. Surrounding farmland has been
damaged by heavy metals. According to measurements of the soil,
the average level of cadmium was up to 60 times the background
value. Grain planted adjacent to the site was inedible because of its
remarkably high content of cadmium.  Therefore, the smelter had
to be closed  down, and  measures dealing with the residues  are
under investigation.
   Moreover,  considerable  amounts  of  hazardous  and  toxic
substances such as arsenic,  chromium, mercury, etc. are washed
away without control into  the  environment, giving rise to con-
tamination problems.
   Self-ignition  of  coal  gangue  is   also a  serious  problem.
Several tons of sulfur dioxide are generated by the combustion and
emitted into the air each day.

MANAGEMENT OF  SOLID WASTES

   It is policy  in China that all types of  industrial wastes should be
reused  or be detoxified prior  to their  disposal.  Especially when
dealing with hazardous wastes, handling shall be extremely careful
to avoid secondary  pollution;  when handling radioactive wastes,
volume  reduction and solidification  shall  be  required before
ultimate disposal.
  On the other hand, the policy of the Government  encourages all
technologies such  as environmental pollution abatement, package,
transfer, storage and destruction of hazardous chemicals, and ad-
vanced industrial processes that produce less pollution or are even
pollution-free.
Comprehensive Utilization

  Such practices are rather attractive because many of the wastes
or used  materials could be reused directly or after simple process-
ing.  There are reclamation  networks available in  most cities in
China to collect scrap metals, waste paper, plastics, rubbers and
textile. Annually,  they turn  over 4 billion  yuan ($1.7 billion, US)
and have net  profits of 0.2 billion yuan ($86 million, US).
  Coal gangue discarded yearly amounts to 100 million tons. The
prediction is that,  in 1985, it  will be 130-180 million tons/yr. Up to
1 billion tons have been accumulated over  the past years. Possible
uses include: burning,  or utilization in building materials (cement,
lime, brick), or using as filler for land reclamation or  road making.1
About 20% of the coal gangue is now being used.
  Eighty percent of the slag has been utilized as a raw material in
cement at an annual coal savings of 2 million tons and cost saving
of about 100  million yuan ($43 million, US)/yr.
  Over ten power stations in China have used all their own fly ash.
In 1982, 12% of all fly ash produced was  utilized by the building
material industry.  Fly ash generation in Shanghai amounts to about
1 million tons/yr, of which nearly 70%  has been beneficially used.
  Another use of fly ash is  in amelioration of soils; experiments
have been ongoing for  several years. If fly ash is applied at a rate of
75 to 300 tons/ha, 12-38% and even over 100% increase in produc-
tion of crops of wheat, paddy, soybean, rape seed, etc., can be  ex-
pected without any adverse effect of harmful substances.'
  Tests  are being carried out in an attempt to determine beneficial
uses for other waste residues. For instance, chromic  wastes may be
used as coloring  matter  for glassware or as  fillers in building
materials after being detoxified. Zhuzhou Smeltery in Hunan Pro-
vince has achieved  an annual  profit  of  10  million yuan  ($4.3
million, US) from recovery of gold, silver  and other precious and
rare metals out of its own waste residues.
604
INTERNATIONAL ACTIVITIES

-------
Disposal Practices
Chromic Wastes
  Over 2 million tons of chromic residues have accumulated in
China. Since enterprises generating the wastes are spread all over
the country, disposal is a difficult task. Currently, a requirement
that chromic residues be detoxified and/or dewatered before fur-
ther use or ultimate disposal is being considered. For instance,
chromic wastes could be melted at a high temperature to form a
kind of glass or be reduceds from Cr(VI) to Cr(III).
  In Jinzhou Ferroalloy Mill of Liaoning Province, there was, at
one time,  a  0.2  million ton dump of  0.5%  Cr(VI)-containing
residues. This 4 ha open dump had caused groundwater contamina-
tion in an area 12.5 km in length and 1 km in breadth. Within this
area, water from 1800 wells was found to be unsuitable for drink-
ing. Attempts had  been  made to  dig  some interceptor  wells
downstream from the dump to collect  and treat contaminated
groundwater. However, those steps failed.1
  Therefore, a remediation project was begun with an investment
of 4.2 million yuan ( $1.8 million, US) to construct a concrete bar-
rier for  pollution control. The 800 m  long  barrier  penetrates
through the earth down to waterproof rock at a maximum depth of
18  m; it rises 2 m above the ground, circling the site. That shape
forms an underground "tank" consisting of rocky bottom and im-
permeable enclosure. The next step is to regulate the groundwater
level within the "tank" to  hinder the leachate  from seeping—by
keeping the level in the tank lower than outside it. Downstream, the
Cr(VI) containing groundwater is pumped out regularly and treated
before discharge.  The  project was completed  in 1982. Nearly  3
months later, Cr(VI) content in the water from the interceptor wells
outside the enclosure dropped sharply to 1 to 5% of its prior level,
while that from the inside rose several-fold.4'5 These data indicate
that the barrier, to some extent, is effective in pollution control.
  Pujiang Chemicals Factory in Shanghai also generates a lot of
chromic  residues. As a consequence of  open  dumping, its sur-
roundings and groundwater  have been  damaged. Thus,  a new
disposal site is under design, and a leachate recycling plan is being
prepared.
  The  proposal is that a 1.5 m high barrier will be constructed to
divide the site into several cells 20 m x  4 m each. The floor and
barrier will be of concrete with a pitch liner. On the floor, there will
be  a series of collection lines to direct the leachate into  a tank
equipped with pumps to recover the liquid for reuse. Although the
system was designed for pollution control via leachate treatment, it
also has the advantage of resource reclamation.

Tailings
  Tailings from Sanshandao Gold Ore-dressing Mill in Shandong
Province contain  cyanides and  other  harmful  substances. To
eliminate the hazard from seepages, a tailings landfill site has been
proposed in the form of an enclosure with each cell 400 m x 400 m
in size and 1,000,000 m3 in volume. Its enclosing wall and bottom
are overlain with a waterproof plastic liner.  In practice, it is possi-
ble to construct a system that will not leak, but if an unfavorable
incident occurs, deposited fine tailings would seal off any fissure to
stop seepage.
  Around the site, a number of interceptor wells are arranged to
monitor seepage,  or in case of leak to avoid a further spread  of
pollution by pumping  out  the  contaminated groundwater.  That
system can be referred to as a kind of hydraulic curtain.
  A windbreak will be installed to control dust. Once the site has
been filled, it will then be covered with rock, grit and earth  and be
vegetated. To minimize the quantity of wastewater to be treated,
some of the leachate will be collected for reuse. An emergency tank
will be installed to receive any overflows of  slurry from the tailing
conveying system.4

Red Mud
  Impermeable material is one key in disposal technology.  A new
kind of red mud compounded  material has been developed  in
China to provide for seepage prevention at red mud disposal sites.4
It has been determined that  76% of the red  mud compounded
material has a coefficient of permeability of approximately 10 ~9
cm/sec,  freeze-  and  alkali-proof  properties  and  satisfactory
strength to meet the specification required for liner materials in safe
landfill disposal.

Radioactive Materials
  The general procedure for disposal of radioactive wastes used to
involve dewatering or incineration followed by sealing with cement
and then deposition in a cave which had been specially designed as
an  ultimate disposal site. That was unlikely to be a satisfactory
long-term  solution.  Recently, new storage  sites  for  both  in-
termediate and permanent purposes are under consideration.
  Generally, radioactive wastes can be  classified as high, medium,
or low activity, and can be further sorted into flammable or inflam-
mable categories. In agreement with these  classifications, the
wastes are to be disposed of in different  ways.  For municipal
readioactive wastes in Shanghai (from laboratories, hospitals, etc.),
an experimental treatment station is being planned. Apart from its
task of handling radioactive waste, the station will also be responsi-
ble  for  development  of  methodologies  and technologies  of
disposal.
  A new incinerator, designed for municipal flammable radioactive
waste disposal, is just ready to go into operation. According to its
design specifications, it has a treatment capacity of 9 tons/month,
beyond the amount of waste generated in Shanghai,  and a scrub-
bing efficiency of 99.99%.
  For the inflammable wastes (e.g., contaminated apparatus, in-
struments, etc.), it is recommended that they be cleaned prior to
storage, if possible. A design for a radioactive waste storage facility
has been completed, and the  project will soon be under construc-
tion. It will have a storage capacity of 15 years for the radioactive
wastes produced  in Shanghai. After being  compacted and then
packed, the wastes are loaded in steel containers in accordance with
their activities and stored in the chambers until ultimate disposal is
required.
CONCLUSIONS
  According to Chapter 3 of "The Environmental Protection Law
(For Trial Implementation)", "discharge of all kinds of harmful
substances shall be in compliance with the criteria set down by the
State".
  Today in China, there is regulated transport for over  6,000
hazardous materials, but there are no particular rules for hazardous
wastes.  At  present, to perfect the  management framework in
China, a series of regulations for the prevention of pollution and a
series of criteria  and  standards for control objectives  are  being
drawn up,  e.g.,  criteria dealing  with residues  and sludges for
agricultural  use, for the building material industry, standards for
control of solid wastes from non-ferrous metallic industries, etc.
  Safe solid hazardous waste disposal is a pressing issue in China.
Disposal technologies, as well as their related techniques, must be
investigated and developed.
REFERENCES
1. Shi Qing Chongqing Environmental Protection, 1, 1984, 5-8.
2. Qin Zhigang, Chongqing Environmental Protection, 2, 1984,
   39.
3. East China Testing Institute of Electric Power, "New Approach
   of Fly Ash  Reclamation in Shanghai,"  First Symposium of
   Solid  Wastes, National Network of Scientific Information of
   Environmental Protection, China, 1983.
4. Wu Junqing,  et  al.,  "Safe  Landfill  Disposal  of  Industrial
   Solid  Wastes," Report No. 5, Symposium of Control Stan-
   dards  of Solid Wastes from  Non-ferrous Metallic Industries,
   China, Sept. 1984.
5. Peng Hui, Environ.  Eng., 3, 1983, 71.
                                                                                     INTERNATIONAL ACTIVITIES
                                                         605

-------
                                                    AUTHOR  INDEX
 Absalon, J.R., '80-53
 Adamowski, S.J., '83-346
 Adams, W.M., '85-108
 Adams, W.R., Jr.. '82-377; '83-352
 Adkins, L.C.. '80-233
 Ahlert, R.C., '82-203; '83-217
 Ahnell, C.P., Jr.. '80-233
 Ainsworth, J.B., '85-185
 Albrecht, O.W.,  '87-248, 393
 Aldis, H., '83-43
 Aldous, K., '50-212
 Alexander, W.J.,  '82-107
 AUcott. G.A., '81-263
 Allen, H.L., '87-110
 Amster, M.B., '83-99
 Anderson, D.A., '83-154
 Anderson, D.C., '81-223
 Appier,  D.A., '82-363
 Arland, F.J., '85-175
 Arlotta, S.V., Jr.. '85-191
 Assink,  J.W., '82-442
 Astle, A.D., '82-326
 Atwell, J.S., '85-352
 Ayres, J.E., '87-359
 Badalamenti. S.. '83-202. 358
 BaUey, P.E., '82^64
 Bailey, T.E.. '82-428
 Bailey, W.A.. '83-449
 Balfour. W.D.,  '82-334
 Ballif, J.D., '82-414
 Barbara, M.A..  '85-237; '85-310
 Bareis, D.L., '83-290
 Barker, L.J., '52-183
 Barkley, N., '82-146
 Barrett, K.W.. '81-14
 Bartolomeo, A.S., '52-156
 Baughman. K.J., '52-58
 Beam. P.M.. '57-84; '55-71
 Beck, W.W., Jr., '50-135; '52-94; '55-13
 Becker. J.C., '85^42
 Beckert, W.F., '52-45
 Beckett, M.J., '52-431
 Beers.  R.H., '5M58
 Beilke, P.J., '52-424
 Bell, R.M., '52-183, 448
 Berk, E., '55-386
 Benson, B.E., '50-91
 Benson, R., '57-84
 Benson, R.C., '50-59; '52-17; '55-71
 Berger, I.S.,  '52-23
Berkowitz,  J., '55-301
Bernard. H.. '50-220
Bilello, L.J., '55-248
Bixler, B.,  '52-141
 Blackman, W.C., Jr.. '50-91
 Blasland. W.V.. Jr..  '57-215; '55-123
 Boa, J.A., Jr., '52-220
 Bogue. R.W., '50-111
 Bond, F.W., '52-118
 Bouck, W.H.,  '57-215
 Boutwell, S.H., '55-135
 Bowders, J.J., '57-165
 Bracken, B.D.. '52-284
 Bradford, M.L.,  '52-299
 Bradshaw, A.D..  '52-183
 Brandwein. D.I.,  '50-262; '57-398
 Brandwein, S.S.,  '82-91
 Brannaka, L.K.,  '87-143
 Brodd, A.R., '82-268
 Brown, K.W., '87-223
 Brown, M.J., '82-363
 Brown, S.M.. '87-79; '55-135
 Brugger. J.E., '80-119. 208; '87-285; '82-12
 Brunsing, T.P.,  '52-249
 Bruehl, D.H., '50-78
 Brunotts, V.A.,  '55-209
 Bryson, H.C.. '80-202
 Buecker. D.A.. '82-299
 Buller. J., '85-395
 Burgess, A.S., '85-331
 Burgher, B., '82-357
 Burns, H., '85-428
 Burrus, B.C., '82-274
 Bush, B., '80-212
 Butler, H.P., '52-418
 Butterfield, W.S.,  '82-52
 Byrd. J.F., '80-1
 Caldwell. S.. '87-14
 Cane, B.H.,  '82-474
 Carter, T.D., '85-63
 Casteel, D.. '80-275
 Celender, J.A., '52-346
 Chaconas, J.T., '57-212
 Chan, R., '55-98
 Chang. S.S..  '57-14
Chase, D.S.,  '55-79
Childs, K.A., '52-437
Cho,  Y., '85^20
Christofano, E.E., '80-107
Christopher, M.T., '80-233
 Chung, N.K., -80-78
 Cibulskis, R.W.,  '82-36
Cichowicz, N.L., '80-239
Clarke, J.H., '85-296
Clay, P.P., '87-45; '82-40; '85-100
Cochran, S.R.,  '82-131
Cochran, S.R.,  Jr.. '80-233
Cohen, S.A., '87-405
 Cole, C.R., '87-306; '82-118
 Collins, J.P.,  '87-2; '85-326
 Collins. L.O., '85-398
 Colonna, R.,  '80-30
 Cook, O.K., '87-63
 Cook. L.R.. '85-280
 Cooper, C.. '87-185
 Cooper. E.W.. '85-338
 Cooper. J.W., '82-244
 Corbett. C.R.. '80-6; '87-5
 Corbo. P., '82-203
 Corn, M.R., '87-70
 Cornaby, B.W.. '82-380
 Cox. G.V..  '87-1
 Cox, R.D..  '82-58. 334

 Dahl, T.O.. '87-329
 Daigler, J.,  '85-296
 Dalton, T.F..  '87-371
 Davey, J.R.. '80-257
 Dawson, G.W., '87-79;  '82-386;  '85-453
 Day, A.R..  '85-140
 Dehn, W.T., '85-313
 Demmy,  R.H.. '87-42
 Devary. J.L.,  '85-117
 DiDomenico. D.,  '82-295
 Diedidue. A.M., '82-354; '85-386
 Diesl, W.F.. '80-78
 Dime. R.A., '85-301
 DiNapoli, J.J.. '82-150
 DiNitto, R.G., '82-111;  '85-130
 DiPuccio. A.,  '82-311
 Dowiak. M.J., '80-131;  '82-187
 Doyle. R.C., '82-209
 Doyle. T.J.. '80-152
 Drake, B., '82-350
 Driscoll, K.H.. '87-103
 Duff, B.M., '82-31
 Duffala. D.S., '82-289
 Duffee, R.A.,  '82-326
 Duke, K.M., '82-380
 Duncan, D., '87-21
 Duvel, W.A.,  '82-86
Dybevick, M.H., '85-248

Earp, R.F.,  '82-58
Eastman, K.W., '85-291
Eimutis, E.C., '87-123
Elkus, B., '82-366
Ellis, R.A.,  '82-340
Eltgroth, M.W., '85-293
Emerson, L.R., '85-209
 Emig, O.K., '82-128
Emrich, G.H., '80-135
 English, C.J.,  '85-453
 Ess, T., '52-390, 408
606        AUTHOR INDEX

-------
Ess, T.H., '57-230
Evans, J.C., '52-175
Evans, R.B.,  '82-11;  '55-28
Everett,  L.O., '82-100

Falcone, J.C., Jr., '82-231
Fang, H-Y, '52-175
Farrell, R.S., '55-140
Farro, A., '83-413
Feld, R.H., '83-68
Fell, G.M., '83-383
Fellows, C.R., '55-37
Ferguson, T., '50-255
Figueroa, E.A., '57-313
Finkel, A.M., '57-341
Fischer, K.E., '50-91
Forrester, R., '5/-326
Fortin, R.L.,  '52-280
Francingues, N.R., '52-220
Franconeri, P., '57-89
Frank, U., '50-165; '57-96, 110
Freed, J.R.,  '50-233
Freestone, F.J., '50-160, 208;  '57-285
Freudenthal, H.G., '52-346
Friedrich, W., '55-169
Furman, C., '52-131

 Gallagher, G.A., '50-85
 Galuzzi, P.,  '52-81
 Garlauskas, A.B., '55-63
 Gay, F.T., III, '52-414
 Geiselman, J.N., '55-266
 Gemmill, D., '55-386
 Geraghty, J.J., '50-49
 Ghassemi, M., '50-160
 Gibbs, L.M., '55-392
 Gilbert, J.M., '52-274
 Gilbertson, M.A., '52-228
 Gillen, B.D.,  '52-27; '55-237
 Gillespie, D.P., '50-125; '57-248
 Glaccum, R.A., '50-59; '57-84
 Goggin,  B., '57-411
 Gold, M.E., '57-387
 Goldman, R.K., '57-215
 Goldstein, P., '55-313
Goliber, P.,  '50-71
Golob, R.S., '57-341
Goltz, R.,  '52-262
Goltz, R.D., '55-202
Goode, D.J.,  '55-161
Gorton, J.C., Jr., '57-10
Goss, L.B., '52-380
Graybill, L., '55-275
Green, J.,  '57-223
Grube, W.E., Jr., '52-191, 249
Gruenfeld, M., '50-165; '57-96; '52-36
Guerrero, P.,  '55-453
Gurka, D.F.,  '52-45
Gushue, J.J.,  '57-359

Haeberer, A.F., '52-45
Hager, D.G.,  '52-259
Hagger, C., '52-45
Haji-Djafari, S., '55-231
Hale, F.D., '55-195
Hammond, J.W.,  '50-250; '57-294
Hanley, M.M., '52-111
Hansel, M.J., '55-253
Hanson, B., '52-141
Hanson, J.B., '57-198
Hardy, U.Z., '50-91
Harman, H.D., Jr., '52-97
Harrington, W.H., '50-107
Harris, D.J., '57-322
Harris, M.R., '55-253
Hartsfield, B., '52-295
Hass, H., '55-169
Hatayama, H.K., '57-149
 Hawkins, C., -'55-395
 Heare, S., '55-395
 Heeb, M., '57-7
 Hemsley, W.T., '50-141
 Henningson, J.C., '55-21
 Hess, J.W.,  '55-108
 Hijazi, N., '55-98
 Hilker, D., '50-212
 Hill, R., '52-233
 Hill, R.D., '50-173
 Hillenbrand, E., '52-357, 461
 Hina, C.E.,  '55-63
 Hines, J.M., '57-70
 Hinrichs, R., '50-71
 Hitchcock, S., '52-97
 Hjersted, N.B., '50-255
 Holberger, R.L.,  '52-451
 Hooper,  M.W., '55-266
 Hopkins, F., '50-255
 Home, A., '57-393
 Horton,  K.A., '57-158
 Housman, J., '50-25
 Housman, J.J., Jr.,  '57-398
 Houston, R.C., '50-224
 Howe, R.W., '52-340
 Hoylman, E.W.,  '52-100
 Hunt, G.E.,  '50-202
 Hupp, W.H., '57-30
 Hwang,  J.C., '57-317

 Ingersoll, T.G., '57-405
 Isaacson, L., '57-158
 Isbister,  J.D., '52-209


 Jacobs, J.H., '52-165
 Jacot, B.J., '55-76
 James, S.C., '50-184; '57-171, 288; '52-70, 131
 Janis, J.R.,  '57-405;  '52-354
 Janisz, A.J.,  '52-52
 Jerrick, N.J., '55-389
 Jhaveri,  V.,  '55-242
 Johnson-Ballard,  J., '57-30
 Johnson, M.G., '57-154
 Johnston, R.H., '55-145
 Jones, A.K.,  '52-183, 448
 Jones, K.H., '52-63
 Jones, R.D., '55-123, 346
 Jones, S.G.,  '55-154
 Jordan, B.H., '52-354

 Kadish, J., '52-458
 Kaplan,  J., '52-131
 Kaschak, W.M., '52-124
 Keitz, E.L., '52-214
 Kennedy, S.M., '57-248
 Kerfoot, W.B., '57-351
 Khan, A.Q.,  '50-226
 Kilpatrick, M., '50-30
 Kim, C.S., '50-212
 Kimball, C.S., '55-68
 Knowles, G.D., '55-346
 Knox, R.C.,  '55-179
 Koerner, R.M., '50-119;  '57-165, 317; '52-12;
   '55-175
 Kopsick, D.A., '52-7
Kosson, D.S., '55-217
Koster, W.C., '50-141
Koutsandreas, J.D.,  '55-449
Kufs, C., '50-30; '52-146
Kuykendall, R.G., '55-459

LaBrecque, D., '55-28
LaFornara, J.P., '57-110, 294
LaGrega, M.D., '57-42
LaMarre, B.L., '52-291
Langner, G.,  '52-141
Larson, R.J.,  '50-180
Lataille,  M.,  '52-57
 Lawson, J.T., '52-474
 LeClare, P.C., '55-398
 Lederman, P.B.,  '50-250; '57-294
 Lee, C.C., '52-214
 Lee, G.W., Jr., '55-123, 346
 Leighty, D.A., '55-79
 Leis, W.M., '50-116
 Leo, J., '52-268
 Lippe, J.C., '55-423
 Lippitt, J.M.,  '52-311; '55-376
 Lipsky, D., '52-81
 Lo, T.Y.R., '55-135
 Lord, A.E., Jr., '50-119, '57-165; '52-12; '55-
   175
 Losche, R., '57-96
 Lough, C.J.,  '52-228
 Loven, C.G., '52-259
 Lowrance,  S.K., '55-1
 Lucas, R.A.,  '52-187
 Lueckel, E.B., '55-326
 Lundy, D.A.,  '52-136
 Lunney, P., '52-70
 Lynch, E.R.,  '57-215
 Lynch, J.W.,  '50-42
 Lysyj, I., '57-114; '55-446
 MacRoberts, P .B.,  '52-289
 Mahan, J.S.,  '52-136
 Malone, P.O., '50-180; '52-220
 Mandel, R.M., '50-21
 Manko, J.M., '57-387
 Martin, W.F., '55-322
 Martin, W.J., '52-198
 Maslansky, S.P., '52-319
 Maslia, M.L., '55-145
 Massey, T.I.,  '50-250
 Mateo, M., '55-413
 Mathamel, M.S., '57-280
 Matthews, R.T.,  '55-362
 Mavraganis, P.J., '55-449
 Mazzacca, A.J., '55-242
 McCloskey, M.H., '52-372
 McCord, A.T., '57-129
 McEnery, C.L., '52-306
 McGarry, F.J., '52-291
 McGinnis, J.T., '52-380
 McKown, G.L., '57-300, 306
 McLaughlin, D.B., '50-66
 McMillion, L.G., '52-100
 McNeill, J.D., '52-1
 Mehran, M., '55-94
 Meier, E.P., '52-45
 Melvold, R.W.,  '57-269
 Menke, J.L., '50-147
 Mercer, J.W., '52-159
 Messick, J.V., '57-263
 Meyer, J., '50-275
 Meyers, T.E., '50-180
 Milbrath, L.W., '57-415
 Miller, D.G., Jr., '52-107;  '55-221
 Montgomery, V.J., '55-8
 Moore, S.F., '50-66
Morahan, T.J., '55-310
Moran, B.V., '55-17
Morey, R.M.,  '57-158
Morgan,  C.H., '50-202
Morgan,  R.C.,  '52-366
Mott, R.M., '50-269; '55-433
Mousa, J.J., '55-86
Muller, B.W.,  '52-268
Muller, Kirchenbauer, H.,  '55-169
Murphy, B.L., '52-331, 396; '55-13
Murphy, C.B., Jr.,  '55-195
Mutch, R.D., Jr., '55-296
Myers, V.B., '52-295; '55-354

Nadeau, P.P., '52-124; '55-313
                                                                                                              AUTHOR INDEX
                                                                                                                                          607

-------
Nagle, E., '83-310
Narang, R.,  '50-212
Nazar, A., '82-187
Neely, N.S., '80-125
Nelson, A.B.,  '87-52
Neumann, C., '82-350
Nielson, M.. '87-374
Niemele,  V.E., '82-437
Nimmons, M.J., '83-94
Nisbet, I.C.T., '82-406
Noel, M.R., '85-71
Noel, I.E., '83-266
Norman, W.R., '82-111
North, B.E..  '87-103
Nygaard, D.D., '83-79

O'Dea, D., '83-331
Ogg, R.N., '83-202, 358
Oi, A.W., '87-122
 O'Keefe, P .,  '80-212
 Openshaw, L-A, '83-326
 Opitz, B.E., '82-198
 Osborn, J.,  '83-43
 Osheka, J.W., '80-184
 Ounanian, D.W., '83-270
 Owens, D.W..  '80-212

 Paige, S.F., '80-30, 202
 Palombo, D.A., '82-165
 Pajak, A.P., '80-184;  '8;-288
 Parks, G.A., '83-280
 Parker, F.L.,  '87-313
 Parratt, R.S.,  '83-195
 Parry, G.D.R., '82-448
 Pearce, R.B.,  '87-255;  '83-320
 Pease, R.W., Jr., '80-147; '87-171, 198
 Peters, J.A.,  '87-123
 Peters, W.R.,  '82-31
 Phillips, J.W.,  '87-206
Pintenich, J.L., '87-70
Possidento, M., '80-25
Possin, B.N., '83-114
Powell, D.H.,  '83-86
Price. D.R., '82-94

Quan, W., '87-380
Quinlivan, S.,  '80-160
Quimby,  J.M., '82-36

Rams.  J.M., '87-21
Ramsey, W.L.. '80-259; '87-212
Ransom,  M., '80-275
Rappaport, A., '87-411
Rebis. E.N., '83-209
Reifsnyder. R.H., '82-237
Reiter, G.A.. '80-21
Remeta, D.P., '80-165; -87-96
Repa, E., '82-146
Richards, A., '80-212
Rikleen, L.S..  '82-470
Ritthaler, W.E., '82-254
Riner, S.D..  '82-228
Rishel, H.L., '87-248
Rizzo, J., '82-17
Robbins,  J.C., '83-431
Roberts, B.R.. '83-135
Rodricks, J.V., '83-401
Rogoshewski, P.J., '80-202; '82-131,  146
Roos, K.S., '83-285
Rosenkranz,  W., '87-7
Rothman, T., '82-363
Roy, A.J., '83-209
Royer,  M.D., '87-269
Rulkens, W.H., '82-442
Ryan, F.B., '8/-10

Sadat, M.M., '83-301, 413
                                        Sanders, D.E.,  '82-461
                                        Sandness, G.A., '87-300; '83-68
                                        Sanning, D.E.,  '87-201; '82-118, 386
                                        Schalla, R., '83-117
                                        Schauf, F.J., '80-125
                                        Schlossnagle, G.W.,  '83-5, 304
                                        Schmidt, C.E.,  '82-334; '83-293
                                        Schnabel, G.A., '80-107
                                        Schneider, P., '80-282
                                        Schneider, R., '80-71
                                        Schoenberger, R.J., '82-156
                                        Schomaker. N.B.,  '80-173; '82-233
                                        Schuller, R.M., '82-94
                                        Schultz, D.W.,  '82-244
                                        Schweitzer, G.E.,  '87-238; '82-399
                                        Scofield, P.A.,  '83-285
                                        Scott, J.C., '87-255;  '83-320
                                        Scott. M., '82-311; '83-376
                                        Scrudato, R.J., '80-71
                                        Seanor. A.M., '87-143
                                        Selig. E.I., '82-458; '83-437
                                        Sevee, J.E.. '82-280
                                        Sewell, G.H., '82-76
                                        Seymour.  R.A.. '82-107
                                        Sharma, O.K., '8/-I85
                                        Shaw, L.G., '8/-415
                                        Sheedy, K.A., '80-116
                                        Shen, T.T.. '82-70, 76
                                        Sherman, J.S., '82-372
                                        Sherwood, D.R., '82-198
                                        Shih. C.S., '8/-230; '82-390,  408; '83-405
                                        Shroads, A.L.. '83-86
                                        Shuckrow, A.J.,  '80-184; '87-288
                                        Shultz, D.W., '82-31
                                        Silbermann. P.T.,  '80-192
                                        Silcox, M.F.,  '83-8
                                        Silka. L.R., '80-45; '82-159
                                        Sills, M.A., '80-192
                                        Simcoe, B., '8/-21
                                        Sims, R.C., '83-226
                                        Singh. R., '83-147
                                        Sirota, E.B., '83-94
                                        Siscanaw, R.,  '82-57
                                        Slack, J., '80-212
                                        Slater, C.S., '82-203
                                        Smith. E.T., '80-212
                                        Smith, M.A..  '82-431
                                        Smith, R., '80-212
                                        Snyder, A.J..  '87-359
                                        Snyder. M., '80-255
                                        Solyom. P., '83-342
                                        Spear, R., '87-89
                                        Spencer, R.W., '82-237
                                        Spittler, T.M., '87-122;  '82-40, 57; '83-100, 105
                                        Spooner, P.A.. '80-30, 202;  '82-191
                                        Springer, C., '82-70
                                        Srivastava, V.K.. '83-231
                                        Stammler, M., '83-68
                                        Stanford, R.,  '87-198
                                        Stankunas, A.F., '82-326
                                        Stanley. E.G.. '83-1
                                        Starr. R.C.. '80-53
                                        St. Glair, A.E.,  '82-372
                                        Steimle, R.R.. '8/-212
                                        Stephens, R.D.,  '80-15; '82-428
                                        Sticf. K.. '82-434
                                        Stoller. P.J.. '80-239; '87-198
                                        Stone. W.L..  '87-188
                                        Strattan, L.W.,  '8/-103
                                        Strauss, J.B., '87-136
                                        Stroud. F.B.,  '82-274
                                        Struzziery. J.J., '80-192
                                        Sullivan,  D.A.,  '87-136
                                        Sullivan, J.H.. '83-37
                                        Swenson, G.A., III, '83-123

                                        Tackett, K.M., '87-123
                                        Tafuri, A.N.. '8/-188; '82-169
 Tanzer, M.S.. '87-10
 Tapscott, G., '82-420
 Taylor. B., '83-304
 Teets, R.W., '83-310
 Tewhey. J.D.. '82-280
 Theisen, H.M.. '82-285
 Thibodeaux, L.J., '8270
 Thompson, S.N., '83-331
 Thorsen. J.W., '87^12, 259; '82-156
 Thomas, G.A., '80-226
 Threlfall, D.. '80-131; '82-187
 Titus. S.E.,  '87-177
 Townsend. R.W.,  '82-61
 Tremblay, J.W.. '83-423
 Triegel. E.K.,  '83-270
 Truett, J.B.,  '82-451
 Tuor, N.R., '83-389
 Turoff, B., '80-282
 Turpin. R.D., '87-110, 277; '83-82
 Turner, J.R., '83-17
 Tusa, W.K., '87-2;  '82-27
 Twedell, A.M., '80-233
 Twedell. D.B.. '80-30. 202
 Tyagi, S.. '82-12

 Unites. D.F., '80-25;  '87-398; '83-13
 Unterberg, W.. '87-188
 Urban. N.W., '82-414; '83-5, 304

 Vanderlaan, G.A., '87-348; '82-321; '83-366
 Vandervort, R., '87-263
 Van Ee, J.J., '83-28
 Van Gemert, W.J. Th. '82-442
 Vogel. G.A.. '82-214

 Wagner. K..  '82-169;  '83-226
 Wallace, L.P.. '83-322
 Wallace, J.R., '83-358
 Waller, M.J., '83-147
 Walsh. J.. '82-311
 Walsh, J.F.,  '8.2-63
 Walsh. J.J., '80-125; '87-248; '83-376
 Walther, E.G.. '83-28
 Wardell, J., '87-374
 Weber, D.D., '83-28
 Weist. F.C.,  '83-175
 Weiaer. P.M.. '87-37
 Welks, K.E., '80-147
 Werner. J.D.. '83-370
 Wetzel. R.S.,  '80-30, 202; '82-169, 191
 Wheatcraft. S.W., '83-108
 While, M., '80-275
 White, R.M.,  '82-91
 Whitlock. S.A., '83-86
 Whittaker. K.F..  '82-262
 Wilder, I., '80-173; '82-233
 Wilkinson, R.R., '80-255
 Wilson. D.C., '80-8
 Wilson, L.G., '82-100
 Wine, J., '83-428
 Wolbach, C.D.. '83-54
 Wolf, F.. '83-43
 Wong, J., '87-374
 Wright, A.P., '80-42
 Wuslich. M.G.. '82-224
 Wyeth, R.K., '87-107
 Wyman, J., '83-395
Yaffe, H.J.,  '80-239
Yang, E.J.. '87-393; '83-370
Yezzi, J.J., Jr.,  '87-285
Young, L., '80-275
Young, R.A., '87-52
Yu, K.. '80-160


Ziegler. F.G., '87-70
608
AUTHOR INDEX

-------
                                                     SUBJECT  INDEX
Above Ground Closure, '83-215
Activated Carbon, '81-314; '82-259, 262; '83-
  209, 248, 253, 342
Air Modeling, '82-331
Air Monitoring,'82-67, 268, 299, 306, 331;
  '83-62, 85
    Ambient, '87-280; '83-293
    Emissions, '82-10
    Nitrogen Compounds, '55-100
    Real Time, '53-98
    Sampling Techniques, '82-334
    Two Stage Tube, '83-85
Air Photos,  '80-116
Air Quality, '82-63
    Assessment,  '82-76
Air Stripping, '83-209, 313, 354
Analysis, '82-45
    Metals,  '83-79
    Portable Instruments, '82-36, 40, 57
    Pyrographic,  '87-114
    Screening, '83-86
    Spectrometer, '83-291
Arizona, TCE Contamination, '82-424
ASCE, '81-2
Assessment,  '82-17, 27; '83-37
    Biological, '82-52
    Cold Weather, '82-254
Closure/Post-Closure,  Illinois Perspective,
  '83-549
CMA, '87-1
Coal Tar Cleanup, '83-331
Community Coordinator,  '8/-411
Community Relations (See also Public
  Participation),  '87-405,  415; '82-354
    Health Concerns,  '82-321
    Program, '83-386, 389
Compatibility Testing, '87-110
Composting, Soils, '82-209
Connecticut, Risk Evaluation, '80-25
Containment System Design,  '82-175
Contaminated Soil, '83-226, 231
    Cleanup, '83-354
Contamination, Mapping, '83-71
Contingency Fund, '80-21
Contingency Plan, Massachusetts, '83-420
Contracts, REM/FIT,  '83-313
Cost,  '80-202; '87-248; '83-209
    Above Ground Waste Storage,  '82-228
    Air  Stripping, '83-313
    CERCLA Financed, '83-395
    Cleanup, '82-262;  '83-296, 366, 370
     Cleanup Level, '83-398
     Computer Models, '83-362
     Cover, '82-187
     Effectiveness Evaluation, '82-372
     Estimates, '80-202
     Groundwater Treatment, '83-248, 358
     Health and Safety Impact, '83-376
     Management,  '87-348, 351
     Mathematical Modeling, '87-306, 313
     Mercury Contamination, '82-81
     Methods, '87-79
     Pesticide Plant, '82-7
 Auditing, '87-398

 Barriers, '82-249
     Bentonite, '82-191
     Gelatinous, '82-198
 Bedrock Aquifer, Contaminant Movement,
   '82-111
 Bench Scale Study, '87-288
 Bench Scale Testing, '80-184
 Berlin & Farro, '87-205
 Biodegradation, '82-203
 Bioindicators, '87-185
 Biological Monitoring, '87-238
 Block Displacement Method,  '82-249
 Bromine, Organic, '82-442
 Buried Drums, Sensing, '80-239

 California Superfund Program, '82-428
 Callahan Site, '82-254
 Capping, '83-123, 296
    Cost, '83-370
 Chemical Analysis,  Rapid, '80-165
 Chemical Control, '87-341
 Chemical Oxidation, '83-253
 Chemical Plant, Emergency Removal, '83-338
 Chlorinated, Hydrocarbons, Groundwater
  Monitoring, '82-1
 Chromium Sludge, '80-259
Clay, Leachate Interaction,  '83-154
    Organic Leachate Effect,  '87-223
Cleanup, '80-147, 257
    Assessment Role, '83-389
    BT-Kemi Dumpsite, '83-342
    Case Studies, '83-395
    Coal Tar, '83-331
    Cold Weather, '82-254
    Criteria,  '83-301
    Delays, '83-320
    Drum Site,  '83-354
    Dual Purpose, '83-352
    Extent, '83-433
    Forced, '87-255
    Gilson Site Proposal, '82-289
    Hardin County Brickyard, '82-274
    Level,  '83-398
    Liability Due to Failure, '83-442
    Long Term Effectiveness, '82-434
    Management, '83-370
    PCB, '82-156, 284
    Picillo  Farm, '82-268
    Staged Approach, '82-262
Closure,  '87-259
    Copper Residue Disposal Site, '87-70
    Impoundment,  '83-195
    Leachate Collection, '83-237
    Leachate Monitoring, '82-97
    Minimization, '87-84
    Recovery Documentation, '82-366
    Remedial, '82-118
    Savings Via Negotiation, '82-377
    Treatment System, '87-294
    Water  Recovery System, '82-136
Coventry, RI, '80-239
Cover (See  also caps), '82-183, 187, 448
Cutoff Wall, '83-123, 296
    Chemically Resistant, '83-169, 179, 191
    Cost, '83-362

Damage, Cost Recovery, '87-393
Data Bases, '83-304
Decision  Making, '87-230
Decision  Tree Analysis, '82-408
Decontamination, '80-226
    Waterway,  '83-21
Degradation, TNT Sludge, '83-270
Denney Farm,  '87-326
Department of Defense Program, '82-128
Design, Mathematical Modeling,  '87-306
    Preliminary, '80-202
Detoxification,  '80-192
DIMP, '87-374
Dioxin, '87-322, 326; '83-405
Dispersion of Coefficients, '83-135
Disposal, '87-329
    Above Ground, '83-275
    Commercial Criteria, '82-224
    Computer Cost Model, '83-362
    Liability, '83-431
    Salt  Cavities, '83-266
Documentation, Cost Recovery, '82-366
                                                                                                           SUBJECT INDEX
                                                                                       609

-------
DOD, Hazardous Materials Technical Center,
   '82-363
    Site Cleanup,  '83-326
Down Hole Sensing, '83-108
Drain System, '83-237
Drums, '52-254
    Buried, '52-12
    Handling, '82-169
    Site Cleanup,  '83-354
Electromagnetic
    Induction, '53-28, 68
    Resistivity,  '82-1
    Survey, '80-59;  '82-12
    Waves, '80-119
Emergency Removal, '83-338
Emissions Monitoring, '53-293
Environmental
    Impact, '5/-177
    Risk Analysis,  '52-380
Excavation, '52-331
Exhumation, '52-150
Exposure-Response Analysis, '52-386

Federal/State  Cooperation, '52-420
Fire,  '57-341;  '52-299
FIT
    Contracts, '83-313
    Health and Safety, '80-85
Florida's Remedial Activities, '82-295
Fort  Miller, '8/-215
Fugitive Hydrocarbon Emission Monitoring,
   '8/-123

Gas Chromatograph, '82-57, 58; '83-76
    Portable, '82-36; '83-105
GC/MS, '82-57
Geohydrology, '83-117
Geophysical, '83-68, 71
    Methods, '52-17
    Monitoring, '53-28
    Survey, 'S/-300
    Techniques, '53-130
Geophysics, '57-84;  '52-91
Geotechnical Techniques, '53-130
Geotechnology, Containment System, '52-175
Gilson Road Site, '82-291
Ground Penetrating  Radar, '80-59, 116, 239;
   '8/-158, 300; '83-68
Groundwater
    Alternatives to Pumping, '82-146
    Cleanup, '82-118, 159; '83-354
    Containment, '82-259; '83-169
    Containment Movement,  '82-111
    Contamination,  '81-329, 359; '82-280;
    '83-43, 358
       Liabilities, '83-437
       Mapping, '83-71
       Potential, '50-45
    Flow System, '53-114,  117
    Hydraulic Evaluation,  '53-123
    Investigation, '50-78
    Mathematical Modeling, '57-346
    Metal Finishing Contamination, '53-346
    Microbial Treatment, '53-242
    Migration, '50-71
      Prevention, '53-179,  191
    Modeling, '52-118; '53-135, 140, 145
    Monitoring, '50-53; '52-17,  165
      Interpretation, '52-86
    Pollution Source, '87-317
    Post-Closure Monitoring, '83-446
    Protection, '80-131
    Recovery Cost, '82-136
    Recovery Design, '82-136
    Remedial Plans,  '83-130
    Research Needs,  '83-449
    Restoration, '82-94
    Sampling,  '87-143, 149
                                          TCE Contamination, '82-424
                                          Treatability,  '87-288
                                          Treatment, '80-184;  '82-259; '83-248, 253
                                      Grout, '53-169, 175
                                          Chemistry, '52-220
                                      Grouting, '82-451
                                          Silicates. '82-237

                                      Hazard, Degree, '87-1
                                      Hazardous Materials Storage, Spills,  '82-357
                                      Hazardous Materials Technical Center, '52-363
                                      Hazard Potential, '50-30
                                      Hazard Ranking,  '57-188
                                          Prioritizing, '57-52
                                          System, '57-14; '82-3%
                                          U.S. Navy Sites, '83-326
                                      Hazards, Unknown, '87-371
                                      Health and Safety (See also Safety)
                                          Community Concerns, '82-321
                                          Cost Impact,  '83-376
                                          Guidelines, '83-322
                                          Hazards, '80-233
                                          Plan.  '83-285
                                          Program, '50-85, 91, 107
                                      Heavy Metals, Impoundment Closure, '83-195
                                      High-Pressure Liquid Chromatography, '83-86
                                      Hydrocarbons, Leaks, '52-107
                                      Hydrogeologic Evaluation, '50-49
                                      Hydrogeological Investigation,  '57-45, 359;
                                        '53-346

                                      Identification, '53-63
                                          Reactivity, '83-54
                                      Illinois, Closure/Post Closure,  '83-459
                                      Immobilization, '82-220
                                      Impact Assessment,  '87-70
                                      Impoundment,  '80-45
                                          Closure, '83-195
                                          Leaks.  '53-147
                                          Membrane Retrofit, '82-244
                                      Incineration, '82-214
                                          Sea, '50-224
                                          Mobile, '50-208; '57-285
                                      Inductive Coupled Plasma Spectrometer,  '83-
                                        79
                                      Insurance,  '82-464
                                      Interagency Management Plans, '80-42
                                      Investigation, Hydrogeologic, '82-280

                                      Kriging,  '80-66

                                      Laboratory Management, '87-96
                                      Laboratory, Regulated Access, '87-103
                                      La Bounty Site, '82-118
                                      Lagoons,  '87-129; '82-262
                                      Landfill
                                          Closure, '80-255
                                          Future Problems, '80-220
                                      Leachate
                                          Clay Interaction, '83-154
                                          Collection, '83-237
                                          Effects on Clay, '87-223
                                          Generation Minimization,  '80-135, 141
                                          Migration, '52-437
                                          Minimization, '57-201
                                          Modeling,  '53-135
                                          Monitoring Cost, '82-97
                                          Treatment, '80-141;  '82-203,437; '83-
                                        202, 217
                                      Leak  Detection. '83-94.  147
                                      Legal Aspects, Extent of Cleanup,  '83-433
                                      Legislation, Model Siting Law,  '80-1
                                      Liability, '82-458  , 461,  464, 474
                                          Corporate, '80-262
                                          Disposal,  '83-431
                                          Generator, '87-387
                                          Groundwater Contamination,  '83-437
                                          Inactive Sites, '80-269
    Superfund Cleanup Failure, '83-442
    Trust Fund,  '83-453
Liner
    Breakthrough, '83-161
    Leak Location, '82-31
    Synthetic Membrane, '83-185
Love Canal.  '80-212, 220; '81-415; '82-159, 399

Magnetrometry. '80-59, 116;  '8/-300, '82-12;
   '83-68
Management Plans, New Jersey, '83-413
Massachusetts Contingency Plan, '83-420
Mercury,  '82-81
Metals, '82-183
    Analysis, '83-79
    Detection, '80-239
    Detector, '80-59; '8/-300; '82-12
    Finishing,  '83-346
Microbial Degradation,  '83-217, 231, 242
Migration
    Cut-Off, '82-191
    Prevention, '82-448
Mining Sites, '53-13
Mobile Laboratory, '80-165
Modeling
    Groundwater Treatment, '83-248
    Leachate Migration, '82-437
    Remedial Action. '83-135
Models
    Management Options, '83-362
    Site Assessment. '87-306
Monitoring, Ambient Air, '87-122, 136
Monitoring Well
    Installation. '87-89
    Location, '87-63

National Priority List, Mining Sites.  '83-13
National Resource Damage, '87-393
National Response, '87-5
Negotiating,  '82-377, 470
Neutralization, '83-63
New Jersey, Cleanup Plans, -83-413
Non-Destructive Testing Methods,  '82-12
Odor,  '82-326; '83-98
Old Hardin County Brickyard,  '82-274
Organic Vapor
    Analysis, '83-98
    Field Screening, '83-76
    Leak Detection, '83-94
    Personnel Protection, '87-277
Organics, Emissions, '82-70
Ott/Story, '87-288

Parametric Analysis, '87-313
PCBs, '87-215; '82-156, 284;  '53-21.  326, 366,
  370
    Field Measurement, '83-105
Pennsylvania's Program,  '87-42
Personnel Protection Levels,  '87-277
Pesticides, '82-7
Picillo Farm  Site,  '82-268
Pilot Plant, '87-374
PIRS.  '82-357
Pittston, PA, '80-250
Plant Bioindicators, '87-185
Post-Closure
    Care. '87-259
    Failure,  '83-453
    Groundwater Monitoring,  '83-446
    Monitoring,  '82-187
    Monitoring Research, '83-449
POTW, Leachate Treatment, '83-202
Price Landfill, Remedial Action, '83-358
Prioritization (See also Hazard Ranking), '87-
   188
Public Awareness, '83-383
Public Information Program, '80-282
610
SUBJECT INDEX

-------
Public Participation (See also Community
  Relations) '82-340, 346, 350; '83-363
    Failures, '83-392
Public Policy, Cleanup Level,  '83-398
Pulsed Radio Frequency, '81-165

Quality Control, '82-45

Radioactive Wastes, '81-206
Radon Gas, '82-198
RAMP, '52-124
    Love Canal, '52-159
Ranking System, '81-14
RDX,  '82-209
Reactivity,  Identification, '55-54
Real Estate, Hazardous Waste Implications,
   '52-474
Reclamation, Chromium Sludge, '50-259
Records Management System,  '57-30
Regional Response Team,  '50-6; '82-214
REM Contracts, '55-313
Remedial Action, '52-289
    Case Studies, '52-131
    Design, '50-202
    Florida's Site, '52-295
    Options, '50-131
    Progress Status, '50-125
Remedial Construction, Safety Plans, '53-280
Remedial Design
    Groundwater, '53-123
    Model Based Methodology, '53-135
Remedial Projects, Corps of Engineers, '53-
   17
Remedial Response, U.S. Army's Role, '52-414
Remote Sensing, '50-59, 239; '57-84, 158, 165,
   171
Research
    Post-Closure Monitoring,  '53-449
    USEPA Program, '50-173
Resistivity, '50-239; '57-158; '52-31; '53-28
Resource Recovery, '57-380
Response
    Model, '57-198
    Procedures, '50-111
Reverse Osmosis, '52-203
Risk
    Acceptability, '53-405
    Analysis, '57-230;  '53-37
       Environmental,  '52-380
    Assessment, '57-238; '52-23, 386, 390,
       406, 408;  '53-342
       Air Quality,  '52-63
       Comparative, '53^401
       Modeling, '52-396
    Cleanup Level, '53-398
     Evaluation, '50-25
     Minimization,  '57-84
 Rocky Mountain Arsenal, '57-374; '52-259

 Safety (See also Health  and Safety),  '52-299,
  306
     Cost Impact, '52-311
     Procedures, '57-269
     Remedial Construction, '53-280
     Sampling and Analysis, '57-263
     Training, '52-319
Sample Thief, '57-154
Sampling, '50-91
     Analysis, Safety, '57-263
     Biological, '52-52
     Drums, '57-154
     Screening, '57-103,  107, 114
     Techniques, '57-143, 149
Screening, Spectrometry, '53-291
Security, '53-310
Seismic Refraction,  '50-239
Sensing, Downhole,  '53-108
Settlement Agreements,  '52-470
 Shenango,  '50-233
 Shope's Landfill, Cleanup, '53-296
Silicate Grouts,  '53-175
Silicates, '52-237
Silresim Site, '52-280
Site
     Assessment, '50-59, 91; '53-221
     Discovery, '53-37
     Evaluation, '50-25,  30
     Hazard Rating, '50-30
     Location, '50-116;  '57-52
     Location Methodology, '50-275
Siting, '50-1
 Slurry Trench, '52-191
 Soil, Extraction, '52-442
 Soil  Contamination, '52-399, 442; '53-43
     International Study, '52-431
 Solidification, '57-206
     Silicates, '52-237
     TNT Sludge, '53-270
 Solvent Mining, '53-231
 Spills, Hazardous Materials Storage, '52-357
 Stabilization, '50-192
 Stabilization/Solidification,  '50-180
 State Participation,  '52-418
 State Plans, New Jersey, '53-413
 State Superfund Program, '52-428
 Steam Stripping, '52-289
 Stringfellow, Site, '50-15, 21
 Superfund
     California,  '57-37
     Cleanup Failure Liability, '53-442
     Drinking Water, '53-8
     Implementation, '53-1
     Federal/State Cooperation,  '53-428
     Management, '53-5
     Private Sector Concerns, '57-10
     Programs
       New Jersey, '52-413
       Texas, '53-423
     State/Federal Cooperation, '57-21
     USEPA Research,  '57-7
Surface Sealing, '57-201
Surface Water Management, '50-152
Sweden, Dump Site Cleanup, '53-342
Sweeney, '52-461
Sylvester, Site, '57-359
Synthetic Membrane, Impoundment Retrofit,
   '52-244

TAT, Health and Safety,  '50-85
Technology Evaluation, '52-233
Texas, Superfund Program, '53-423
TNT, '52-209
Top-Sealing, '50-135
Trace Atmospheric Gas Analyzer, '53-98,  100
Training, Resources, '53-304
Treatment
     In Situ, '52-451; '53-217, 221, 226, 231
     On-Site, '52-442
     System Design, '57-294

United Kingdom, '50-8, 226
U.S. Army Corps of Engineers, '52-414; '53-
  17
USCG,  '50-6
USEPA
    Mobile Incinerator, '57-285
    Research, '57-7
Vadose Zone Monitoring, '52-100
Vapor Emission, '52-326
Volatile Nitrogen Compounds, Monitoring,
   '53-100
Volatile Organic Emissions, '57-129
Volatile Organics, Monitoring,  '57-122

Walls
     Gelatinous,  '52-198
     Slurry, '52-191
Waste Storage, Above Ground, '52-228
Wastewater Treatment,  '50-160
Water Treatment, Cost, '53-370
Waterway Decontamination,  '53-21
West Germany, '53-68
Wilsonville, Exhumation,  '52-156
Woburn, MA, '57-63, 177
Wood Treating Facility, '57-212
                                                                                                              SUBJECT INDEX
                                                                                           611

-------
           NATIONAL PRIORITIES  LIST—PROPOSED SITES
                                 UPDATE  #2,  OCTOBER 1984
     (Refer to the Proceedings of the 1983 National Conference on Management of Uncontrolled
                  Hazardous  Waste Sites for the Original NPL and  October 1983 Update)
 ALABAMA (04)
 'Alabama Army Ammunition Plant.
  Childersburg
 •Anniston Army Depot (SE Ind Area),
    Anniston

 ARIZONA (09)
  Motorola, Inc., 52nd St. Plant, Phoenix

 ARKANSAS (06)
  Midland Products, Ola/Birta

 CALIFORNIA (09)
  Advanced Micro Devices, Inc., Sunnyvale
  AJviso Dumping Areas, Alviso
  Applied Materials, Santa Clara
  Beckman Instruments, Porterville
  Fairchild Camera, Mountain View
  Fairchild Camera, S. San Jose Plant, South
    San Jose
  Firestone Tire, Salinas
  FMC Corp., Fresno
  Hewlett Packard, Palo Alto
  Intel Corp., Mountain View
  Intel Corp., Santa Clara III, Santa Clara
  Intel Magnetics, Santa Clara
  IBM Corp- San Jose
  J.H. Baxter Co., Weed
  Lorentz Barrel & Drum Co., San Jose
  Louisiana-Pacific Corp.,Oroville
  Marley Cooling Tower Co., Stockton
  Monolithic Memories, Inc., Sunnyvale
  Monlrose Chemical Corp., Torrance
  National Semiconductor Corp., Santa Clara
  Operating Industries, Inc., Lf, Monterey Park
  Precision Monolithic, Inc., Santa Clara
  Raytheon Corp.,  Mountain View
  San Fernando Valley (Area 1), Los Angeles
  San Fernando Valley (Area 2), Los Angeles/
    Glendale
  San Fernando Valley (Area 3), Glendale
  San Fernando Valley (Area 4), Los Angeles
  Slgnetics, Inc., Sunnyvale
  Southern Pacific Transportation, Roseville
  Teledyne Semiconductor, Mountain View
  Thompson-Hayward Chemical Co., Fresno
  Van  Waters & Rogers, Inc., San Jose
  Westinglhouse, Sunnyvale
  Zoecon Corp/Rhone-Poulenc, Inc., East
    Palo Alto
 •Castle Air Force Base, Merced
 • Federal site
•Lawrence Llvermore Lab (USDOE),
    Livermore
•Mather AFB (AC&W Disposal Site),
    Sacramento
•McClellan AFB (Groundwater Com.),
    Sacramento
•Norton Air Force Base, San Bernardino
•Sacramento Army Depot, Sacramento
•Sharps Army Depot, Lathrop

COLORADO (08)
 Eagle Mine, Minturn/Redcliff
 Smuggler Mountain, Aspen
 Uravan Uranium Project, Uravan
•Rocky Flats Plant (USDOE), Golden
•Rocky Mountain Arsenal, Adams County

DELAWARE
•Dover Air Force Base, Dover

FLORIDA (04)
 City Industries, Inc., Orlando
 Davidson Lumber Co., South Miami
 Dubose Oil Products Co., Cantonment
 Montco Research Products, Inc., Hollister
 Peak Oil Co./Bay Drum Co., Tampa
 Pratt and Whitney Aircraft, West Palm Beach

GEORGIA (04)
•Robins Air Force Base, Houston County

HAWAII (09)
 Kunia Wells I, Oahu
 Kunia Wells II, Oahu
 Mililani Wells. Oahu
 Waiiiwa Shaft, Oahu
 Walpahu Wells, Oahu
 Waiplo Heights Wells II, Oahu

ILLINOIS (05)
 Kerr-McGee (Kress Creek) DuPage County
 Kerr-McGee (Rced-Keppler Park), West
    Chicago
 Kerr-McGee (Residential Areas), West
    Chicago
 Kerr-McGee (Sewage Treat Plant), West
    Chicago
 Ml Industries/Taracop Lead Smelt, Granite
  City
 Pagel's Pit, Rock ford
 Peterson Sand & Gravel, Libertyville
 Sheffield  (U.S. Ecology). Sheffield
•Joliet Army Ammunition Plant, Joliet
 Sangamo Crab Orchard NWR (USDOI),
    Cart em lie
•Savanna Army Depot Activity, Savanna

INDIANA (05)
 Fort Wayne Reduction Dump, Fort Wayne
 International Minerals (E. Plant), Terre Haute
 MIDCO II, Gary
 Neat's Dump (Spencer), Spencer

IOWA (07)
 Cbemplex Co., Clinton
 I'.S. Nameplate Co., Mount Vernon
 Vogel Paint & Wax Co.. Sioux City

KANSAS (07)
 Big River Sand Co., Witchita
 National Industrial Environ Serv, Furley
 Slrother Field Industrial Park, Cowley  County
 Maxey Flats Nuclear Disposal, HilLsboro
 Smith's Farm, Brooks

LOUISIANA (06)
•Louisiana Army Ammunition Plant, Doyline

MAINE (01)
•Brunswick Naval Air Station, Brunswick

MARYLAND (03)
 Kane & Lombard Street Drums, Baltimore
 Mid-Atlantic Wood Preservers, Inc., Harmans
 Southern Maryland Wood Treating,
    Hollywood

MASSACHUSETTS (01)
 Haverhill Municipal Landfill, Haverhill
 Norwood PCBs, Norwood
 Rose Disposal Pit, Lanesboro
 Salem Acres, Salem
 Shpack Landfill, Norton/Attleboro

MICHIGAN (05)
 Avenue "E"  Groundwaler Contamin,
    Traverse City
 E.I. duPont (Montague Plant), Montague
 Lacks Industries, Inc., Grand Rapids
 Lenawee Disposal Service, Inc. Lf, Adrian
 Michigan Disposal (Cork St Lf), Kalamazoo
 Motor Wheel, Inc.,  Lansing
 North Bronson Industrial Area, Bronson
 Roto-Finish Co., Inc., Kalamazoo
612
         NATIONAL PRIORITIES LIST

-------
 South Macomb Disposal (Lf #9), Macomb
    Township
 Thermo-Chem, Inc., Muskegon
 Torch Lake, Houghton County
 Waste Management (Holl Lagoons), Holland
 Adrian Municipal Well Field, Adrian
 Agate Lake Scrapyard, Fairview Township
 Koch Refining Co./N-Ren Corp., Pine Bend
 Kummer Sanitary Landfill, Bemidji
 Kurt Manufacturing Co., Fridley
 Long Prairie Groundwater Contam, Long
    Prairie
 Oak Grove Sanitary Landfill, Oak Grove
    Township
 Olmsted County Sanitary Landfill, Oronoco
 Pine Bend/Crosby American Lf, Dakota
    County
 U of Minnesota Rosemount Research Cent,
    Rosemount
 Windom Dump, Windom

 MISSISSIPPI (04)
 Newsom Brothers/Old Reichhold, Columbia

 MISSOURI (07)
 Bee Cee Manufacturing Co., Maiden
 Findett Corp., St. Charles
 Lee Chemical, Liberty
 North-U Drive Well Contamination, Spring-
    field
 Quality Plating, Sikeston
 Solid State Circuits, Inc., Republic
 'Lake City Army Plant (NW Lagoon),
    Independence
 *Weldon Spring  Quarry (USDOE/Army),
    St. Charles County

 MONTANA (08)
 Burlington Northern (Somers Plant), Somers
 Idaho Pole Co., Bozeman
 Mouat Industries, Columbus

 NEBRASKA (07)
 Hastings Groundwater Contamin, Hastings
 Lindsay Manufacturing Co., Lindsay
 Waverly Ground Water Contamin, Waverly
 *Cornhusker Army Ammunition Plant, Hall
    County

 NEW HAMPSHIRE (01)
 Coakley Landfill, North Hampton

 NEW JERSEY (02)
 Cinnaminson Ground Water Contam,
    Cinnaminson
 Fried Industries, East Brunswick Township
 Glen Ridge Radium Site, Glen Ridge
 Jame Fine Chemical, Bound Brook
 Lodi Municipal Well, Lodi
 Montclair/West Orange Radium Site,
    Montclair/West Orange
 Pomona Oaks Residential Wells, Galloway
    Township
 Waldick Aerospace Devices,  Inc., Wall
    Township
'Fort Dix (Landfill Site), Trenton
*Naval Weapons Stat Earle (Site A), Colts
    Neck

NEW YORK (02)
 Anchor Chemicals, Hicksville
 Applied Environmental Services, Glenwood
    Landing
 Byron Barrel & Drum, Byron
 BEC Trucking, Town of Vestal
 Claremont Polychemical, Old Bethpage
 Clothier Disposal, Town of Granby
 • Federal site
 Colesville Municipal Landfill, Town of
    Colesville
 Cortese Landfill, Village of
    Narrowsburg
 Endicott Village Well Field, Village of
    Endicott
 FMC Corp. (Dublin Road Landfill), Town of
    Shelby
 Goldisc Recordings, Inc., Holbrook
 Haviland Complex, Town of Hyde Park
 Hertel Landfill, Plattekill
 Hooker Chemical/Ruco Polymer Corp,
    Hicksville
 Johnstown City Landfill, Town of Johnstown
 Katonah Municipal Well, Town, of Bedford
 Kenmark Textile Corp., Farmingdale
 Liberty Industrial Finishing, Farmingdale
 Nepera Chemical Co., Inc., Maybrook
 North Sea Municipal Landfill, North Sea
 Pasley Solvents & Chemicals, Inc., Hempstead
 Preferred Plating Corp., Farmingdale
 Robintech, Inc./National Pipe Co., Town of
    Vestal
 Sarney Farm, Amenia
 Suffern Village Well Field, Village of Suffern
 SMS Instruments, Inc., Deer Park
 Tronic Plating, Co., Inc., Farmingdale
 Volney Municipal Landfill, Town of Volney
*Griffiss Air Force Base, Rome

NORTH CAROLINA (04)
 Bypass 601 Groundwater Contam, Concord
 Celanese (Shelby Fiber Operations), Shelby
 Jadco-Hughes Facility, Belmont
 NC State U (Lot 86, Farm Unit #1), Raleigh

OHIO (05)
 Alsco Anaconda, Gnadenhutten
 General Electric (Coshocton Plant),
    Coshocton
 Industrial Excess Landfill, Uniontown
 Republic Steel Corp. Quarry, Elyria
 Sanitary Landfil Co. (IWD), Dayton
 Van Dale Junkyard, Marietta

OREGON (10)
 Martin-Marietta Aluminum, Inc., The Dalles
*Umatilla Army Depot, Hermiston

PENNSYLVANIA (03)
 Ambler Asbestos Piles, Ambler
 Brown's Battery Breaking, Shoemakersville
 Domino Salvage Yard, Valley Township
 Hunterstown Road, Straban Township
 Middletown Air Field, Middletown
 Modern Sanitation Landfill, Lower Windsor
    Township
 Shriver's Corner, Straban Township
 Westinghouse Elevator Co. Plant, Gettysburg
 Whitmoyer Laboratories, Jackson Township
*Letterkenny Army  Depot (SE Area),
    Chambersburg

RHODE ISLAND (01)
 Central Landfill, Johnston

TENNESSEE (04)
 American Creosote Works, Inc., Jackson
*Milan Army Ammunition Plant, Milan

TEXAS (06)
 Bailey Waste Disposal, Bridge City
 Brio Refining Co., Inc., Friendswood
 Crystal City Airport, Crystal City
 Koppers Co., Inc. (Texarkana Plant),
    Texarkana
 North Calvacade Street, Houston
 Odessa Chromium tt\, Odessa
 Odessa Chromium tfi (Andrews Hgwy),
    Odessa
 Pesses Chemical Co., Fort Worth
 Petro-Chemical (Turtle Bayou), Liberty
    County
 Sol Lynn/Industrial Transformers, Houston
 South Cavalcade Street, Houston
 Stewco, Inc., Waskom
*Air Force Plant #4 (Gen Dynamics), Fort
    Worth
*Lone Star Army Ammunition Plant, Tex-
    arkana

UTAH (08);
 Mayflower Mountain Tailings Ponds,
    Wasatch County
 Monticello Rad Contaminated Props,
    Monticello
 Olson/Neihart  Reservoir, Wastach County
 Portland Cement (Kiln Dust 2 & 3), Salt
    Lake City
 Sharon Steel (Mid vale Smelter), Mid vale
*Hill Air Force Base, Ogden
*Odgen Defense Depot, Ogden
*Tooele Army Depot (North Area), Tooele

VIRGINIA (03)
 Avtex Fibers, Inc., Front Royal
 Culpeper Wood Preservers, Inc., Culpeper
 IBM Corp. (Masassas Plant Spill), Manassas
 L.A. Clarke &  Son, Spotsylvania County
 Rhinehart Tire  Fire Dump, Frederick County
*Defense General Supply Center, Chesterfield
    County

WASHINGTON (10)
 Mica Landfill, Mica
 Midway Landfill, Kent
 Northside Landfill, Spokane
 Northwest Transformer, Everson
 Quendall Terminal, Renton
 Silver Mountain Mine, Loomis
 Toftdahl Drums, Brush Prairie
*Bangor Ordnance Disposal, Bremerton
*Fort Lewis (Landfill No. 5), Tacoma
*McChord AFB  (Wash Rack/Treatment),
    Tacoma

WEST VIRGINIA (03)
 Mobay Chemical (New Martinsville), New
    Martinsville
 Ordnance Works Disposal Areas, Morgan-
    town

WISCONSIN (05)
 Fadrowski Drum Disposal, Franklin
 National Presto Industries, Inc., Eau Claire
 Stoughton City Landfill, Stoughton

Total Sites Listed: 244


SITES  CONTINUED TO  BE
      PROPOSED  FROM
          UPDATE #1

GEORGIA (04)
 Olin Corp. (Areas 1, 2 & 4), Augusta

OKLAHOMA (06)
 Sand Springs Petrochemical Cmplx., Sand
    Springs

TEXAS (06)
 Pig Road, New Waverly

MISSOURI (07)
 Quail Run Mobile Manor, Gray Summit
                                                                                               NATIONAL PRIORITIES LIST 613

-------
                                                      1984  EXHIBITORS
ACES
115 Gibraltar Road
Horsham, PA 19044
                                  215/441-5924
Associated Chemical and  Environmental  Services
(ACES) is one of the nation's largest  hazardous
material and waste handling contractors with over
20  years  of experience in  site remediation, spill
cleanup,  and  industrial maintenance.  ACES  has
over 300 clients with projects distributed in both the
public and  private sectors. Work activity takes
ACES to any of the 48 continental United States. In
addition  to  site  cleanup services,  ACES provides
waste  water treatment,  physical/chemical  waste
stabilization,  and  underground  storage   tank
remediation.
AIR-TECH Industries, lac.
Post Office Box 507
East Rutherford, NJ 07073
                      201/460-9730
AIR-TECH Industries, Inc.,  East Rutherford, NJ,
has over 25 years of manufacturing experience in
fabric structure technology and offers a wide range
of air supported structures and tension structures
specifically designed to meet the requirements  of
covering waste sites for removal of contaminated
material. AIR-TECH is now actively participating in
USEPA Superfund projects.
ALERT, INC.
P.O. Box 208
Canton, OH 44701
                      216/454-8304
Independent environmental testing laboratories with
mobile analytical  capabilities.  Headquarters
laboratory offers complete organic and inorganic
analysis of  waste,  water,  etc.  with  overnight
emergency turnaround. Mobile units equipped with
G.C./M.S..  O.C.,  atomic  absorption  and  air
monitors for lab packing, waste screening, landfill
assessment and chemical spills. Experienced Super-
fund and ERCS subcontractor.
Acrei American Incorporated
1000 Liberty Bldg.
424 Main Street
Buffalo, NY 14202                  716/853-7525
Acres American Incorporated, an  internationally
known consulting engineering and project manage-
ment firm, provides services to the solid and hazar-
dous waste industry including: hydrogeological in-
vestigations, groundwater monitoring and evalua-
tions, design of TSDF facilities and remediation
programs, and facility closure planning.
                                        AeroVlronmeat, Inc.
                                        145 Vista Ave.
                                        Pasadena, CA 91107
                                                                                     818/449-4392
                                        Hazardous waste site investigations and toxic air
                                        pollutant studies. Special field studies; simulation
                                        modeling of groundwater and air; air pollutant per-
                                        mit analyses; expert testimony.

                                        Air Pollution Control Association
                                        P.O. Box 2861
                                        Pittsburgh, PA 15230               412/621-1090
                                        Publications  and member services in the areas of
                                        hazardous waste management and air pollution con-
                                        trol.
                                                    American Resource* Corp.
                                                    850 West Valley Forge Road
                                                    King of Prussia. PA 19406
                                                                         215/227-7373
American Resources provides  services  and  pro-
prietary technology for the fixation/solidification of
industrial wastes,  including in situ closure of im-
poundments and design or operation of dedicated
processing  facilities.  The technology  is currently
marketed under the PERMIX™ tradename.
American Technological Industries, Inc.
25 S. Shore Dr., P.O. Box 1726
Toms River, NJ 08754         201/255-5163,5900
American Technological Industries, Inc., is a hazar-
dous waste management firm offering transporta-
tion services,  removing and disposal of industrial
and governmental  waste. Our  customers  include
many of the Fortune 500 companies in the U.S. and
various state  and  federal agencies  involved with
waste removal.
                                        Analytical Instrument Development
                                        Route 41  & Newark Road
                                        Avondale, PA19311
                                  215/268-3181
                                        Analytical Instrument Development manufacturers
                                        portable instrumentation for the determination of
                                        trace organic materials in  the environment.  The
                                        Model 511 portable gas chromatograph with elec-
                                        tron capture detection for PCBs in soil will be on ex-
                                        hibit. In addition AID'S new Model 590 GC/OVM
                                        for total organic vapors (OVM) or specific materials
                                        (GC) in  air will be shown. Other AID instruments
                                        for on-site organic measurements at waste sites will
                                        also be displayed.
                                        Aspen Systems Corporation
                                        1600 Research Blvd.
                                        Rockville, MD 20850               301 /251 -5229
                                        Hazardous Waste Report—Newsletter Service
                                                    BCM EsBtern Inc.
                                                    Plymouth Meeting Mall
                                                    Plymouth Meeting. PA 19462         215/825-3800
                                                                                          ext. 360
                                                    Facilities  design  and  environmental engineering
                                                    specializing in  hazardous waste management and
                                                    remediation.
                                                    Baron-BUkesfet, Inc.
                                                    2001 N. Janice Avenue
                                                    MelrosePark. 1L60160
                                  312/450-3913
                                                    Air stripping lowers for volatile organic compound
                                                    removal from water.
Bergen Band A Dram Co.
43-45 O'Brien Street
Kearny. NJ 07032
                                                                                                                             201/998-3500
                                                                                           Bergen Barrel's "Super Shipper" line of closed head
                                                                                           polyethylene drums is available in a 15 through 55
                                                                                           gallon  size.  Our  open top line of polydrums is
                                                                                           available in a 14 through 55 gallon size. Both lines of
                                                                                           polydrums are DOT approved for hazardous waste.
                                                                                           Btoapherks Incorporated
                                                                                           4928 Wyaconda Road
                                                                                           Rockville, MD 20852
                                                                                                                                         301/770-7700
Biospherics provides quality and responsive analyti-
cal services by state-of-the-art methodologies. The
Laboratory utilizes 25,000 sq. feet of space equipped
wilh GC/MS, GC, AA. TOX, TOC, HPLC. IR,
UV-VIS  autoanalyzer  and   full  wet  chemistry
capabilities.  Biospherics  provides  programs  for
priority-pollutant  analyses;  RCRA,  NPDES, and
NIPDWR compliance;  Industrial Hygiene testing,
sampling and consultation; wastewater  treatment;
trealability and pilot plant studies; aquatic bioassays
and environmental field studies.
                                                    Brtgoll Sponge International, Inc.
                                                    3501 Launcelot Way
                                                    Annadale, VA 22003                703/560-7409
                                                                                          373-3482
                                                    Bregoil Sponge International, Inc., is the manufac-
                                                    turer of Bregoil which is a product processed from
                                                    wood cellulose fibers and has millions of capillary
                                                    traps  or pockets. These hold confined hydrocar-
                                                    bons, oils,  and many chemical wastes while  they
                                                    float and repel  water. Bregoil spread and not re-
                                                    trieved degrades to organic mulch much as a plant's
                                                    root. The saturated kernels of Bregoil may be buried
                                                    or incinerated with negligible ash.
614
1984 EXHIBITORS

-------
The Bureau of National Affairs, Inc.
1231 25th Street, NW
Washington, DC 20037              202/452-4452
Since  1970,  with  the  launch of  Environment
Reporter, BNA's  environmental  protection  and
safety information services have set the standard of
excellence. From the wide perspective of Environ-
ment Reporter to the new International Hazardous
Materials Transport Manual, BNA fills a vital need
for factual information throughout the environment
and safety fields.
CECOS International
2321 Kenmore Avenue
Buffalo, NY 14207
716/873-4200
CECOS International, with corporate headquarters
in Buffalo, NY, is the technological leader in the
chemical and hazardous waste management  field
specializing in the safe disposal and management of
toxic  and  hazardous  wastes.  CECOS'  current
capabilities and services include secure  chemical
management  facilities (SCMF),  wastewater  treat-
ment  facilities,  acid  neutralization, lime reclama-
tion,  fuels  blending,   solidification,  PCB
transformer decommissioning, deep well injection
and several ancillary services.

CH2M HILL, Inc.
1941 Roland Clark Place
Reston, VA 22091                   703/620-5200
Founded in 1946, CH2M HILL is one of the largest
engineering firms in the United States with a staff of
over 2,000 men  and women. We provide com-
prehensive study, design, and construction manage-
ment services for technological systems that include
water,  waste management,  agriculture, energy, in-
dustry, transportation, and  civil engineering.
 California Analytical Labs, Inc.
 2544 Industrial Blvd.
 West Sacramento, CA 95691
916/372-1393
 CAL Labs' staff (4  PhD and 30  BS/MS  level
 chemists) and equipment (7 GC/MS, 15 GC, 3 AA,
 1  ICP) occupy  25,000 sq.  feet of  lab  space in
 Sacramento, CA. CAL Labs  performs analyses of
 hazardous wastes for  federal and state regulatory
 agencies as well as private and industrial clients.
 CAL Labs now operates a  Finnigan 8222  high
 resolution GC/MS system, particularly useful for
 TCDD measurements.

 Camp Dresser & McKee
 One Center Plaza
 Boston, MA 02108                  617/742-5151
 Camp Dresser  & McKee (COM) provides  engineer-
 ing and management services  to public and private
 clients responsible  for managing environmental
 resources, facilities, and infrastructure. CDM has
 conducted site investigations and feasibility studies
 and provided design and construction  management
 services at approximately 100  hazardous waste sites
 in the United States.
Century Laboratories, Inc.
1501 Grandview Ave.
Thorofare, NJ 08086
609/848-3939
Century Laboratories, Inc. is a full service, full 24
hour operation laboratory specializing in all areas of
environmental testing and field service work. The
firm's capabilities include: extensive  experience in
the evaluation of hazardous waste, toxic chemicals,
and soils analysis  utilizing  such  state-of-the-art
analytical tools in priority pollutant analysis, com-
plete  RCRA  testing EP toxicity  analysis,  and
bioassay determinations. To  complete our full ser-
vice capabilities we aid our clients in preparation of
bids, manifests, report writing and data evaluation.
Field service teams for collecting both air and waste
samples  are  available for on-site work throughout
the United States.

Chemfix Technologies, Inc.
Post Office Box 1572
Kenner, LA 70063                   504/467-2800
Chemfix  Technologies,  Inc.,  (CTI)  offers  the
                    patented CHEMFIX8 Process for the chemical fix-
                    ation/stabilization  of both  hazardous and non-
                    hazardous liquid and sludges. Complete mobile pro-
                    ject services are offered,  as well as fixed plant
                    facilities for continuous generation waste streams.
                    CTI services include  site assessment,  waste  stream
                    characterization, and engineering and  permitting
                    support.

                    Chemical Waste Management, Inc.
                      ENRAC Division
                    3003 Butterfield Road
                                                      Oak Brook, IL 60521
                                                                                         312/654-8800
ENRAC serves industry and regulatory agencies in
all aspects of hazardous waste mitigation. ENRAC's
network of  analytical laboratories, specialized on-
site equipment, transportation fleet, and past ex-
perience in all phases of remedial action from site in-
vestigations to surface and sub-surface cleanup, sup-
port  its  capability  to  handle  situations  from
emergency spills, and plant cleanups to  large-scale
remedial action.
                    Chem-Met Services
                    18550 Allen Road
                    Wyandotte, MI 48192
                                                      313/282-9250
                    Chem-Met Services  provides environmentally safe
                    treatment fo liquid and solid wastes. Since 1966 we
                    have  provided  an  ecologically  sound   and
                    economically efficient  process of stabilizing waste
                    streams. As a client-oriented company we enjoy pro-
                    cessing acids,   alkalis,  adhesives,  paints,  oils,
                    phosphates, resins, plastisols, solvents, sludges and
                    other hazardous wastes into an inert solid.
Clayton Environmental Consultants, Inc.
25711 Southfield Road
Southfield, MI 48075                313/424-8860
Clayton Environmental Consultants,  a nationally
recognized consulting  firm, assists in developing
waste management action plans and strategies, in-
cluding:  air,  water,  and  earth  hazardous  con-
taminants studies; accredited and licensed analytical
services;  hydrogeological  investigations;  expert
testimony; environmental risk assessments; inactive
hazardous waste site investigations, and many more.

CompuChem Laboratories
P.O. Box 12652
Research Triangle Park, NC 27709    800/334-8525
CompuChem Laboratories is  the  world's  largest
laboratory specializing in Hazardous Waste Analysis
by  GC/MS.  With its  extensive experience in the
field, CompuChem is  able to  provide a range of
analytical  laboratory services to meet  the needs of
clients in the following areas: Superfund waste site
analysis; RCRA;  priority pollutant analysis; iden-
tification  of  unknown  wastes; groundwater
monitoring; dioxin analysis;
continued  on  attached sheet—DAN-there is no
attached sheet!!

Crown Zellerbach
Nonwoven Fabrics Division
3720 Grant St.
Washougal, WA 98671               800/426-0700
Crown Zellerbach's Fibretex display booth offers
product development literature, specifications, ap-
plication guides and samples of Fibretex geotextile
fabrics. Fibretex  provides protection  to geomem-
branes  in ponds,  landfills  and  reservoirs  by:
cushioning against  surface  abrasion  or  puncture
from  sharp  rocks; providing lateral  venting  of
subgrade gases to avoid floating or rupture.
                    Cytox/Polybac Corporation
                    954 Marcon Blvd.
                    Allentown, PA 18103
                                   215/264-8740
                    Cytox/Polybac Corporation develops and manufac-
                    tures  specialized  bacteria,  enzymes,  specialty
                    chemical formulations and equipment systems  for
                    application in pollution control and agriculture.
                                                      Darell Bevis Associates, Inc.
                                                      Route 2, Box 311
                                                      Sterling, VA 22170                  703/430-7100
                                                      Darell Bevis Associates, Inc. provides services and
                                                      products  in Occupational Respiratory  Protection
                                                      and  Hazardous  Materials  Response.   Training
                                                      courses include: Occupational Respiratory Protec-
                                                      tion and  Protecting HazMat Response  Personnel.
                                                      Consultation  services  include:  expert  witness
                                                      testimony, research studies, fit  testing,  program
                                                      audits,  and custom designed training courses and
                                                      audiovisuals. Also  available are audiovisual pro-
                                                      grams and publications.
                                                                          Dunn Geoscience Corporation
                                                                          5 Northway Lane North
                                                                          Latham, NY 12110
                                   518/783-8102
Dunn   Geoscience  Corporation,   Geologic  and
Hydrologic   Consultants:  hazardous  waste  site
assessment   and  remediation  services  include:
geophysical  surveys; aquifer pump tests; ground-
water   sampling;  groundwater modeling;   com-
puterized data  analysis;  regulatory  compliance
assistance. Offices in Albany, Buffalo, Laconia and
Harrisburg.

EA Engineering, Science, and
  Technology, Inc.
15 Loveton Circle
Sparks, MD 21152                  301/771-4950
EA Engineers and Scientists provide assistance in
complying with environmental regulations and pro-
tection  objectives  of our clients in industry and
government  in the following areas: planning; ap-
plied research; engineering design;  site assessments
and field investigations; environmental engineering;
toxic and hazardous materials management; plan-
ning, regulatory liaison and compliance monitoring;
and environmental risk management.
                                                                          EAL Corporation
                                                                          2030 Wright Avenue
                                                                          Richmond, CA 94804
                                   415/235-2633
                                                                          EAL  Corporation  provides  consulting  and
                                                                          analytical  services  in  the  technical fields of en-
                                                                          vironmental science, occupational health and safety,
                                                                          and nuclear science. EAL's exhibit shows how in-
                                                                          dustry and government agencies, as well as the con-
                                                                          sulting  engineering profession, can  utilize EAL's
                                                                          services to support their hazardous waste programs.
                                                                          ENRECO, Inc.
                                                                          5772 Canyon Drive
                                                                          Amarillo, TX 79109
                                   806/359-3511
                                                                          ENRECO, Inc., uses specially designed equipment
                                                                          to inject and thoroughly mix fixation/solidification
                                                                          chemicals  into  existing   waste  lagoons.  These
                                                                          chemicals convert the liquid or semi-liquid waste in-
                                                                          to a solid soil-like material.  Our process has been
                                                                          used for remedial action as well as for treatment of
                                                                          waste prior to landfilling on site.
                                                                          ENSCO, Inc.
                                                                          1015 Louisiana Street
                                                                          Little Rock, AR 72202
                                   501/375-8444
                                                                          A full PCB service company: high temperature in-
                                                                          cineration of PCBs, solids,  liquids and capacitors;
                                                                          transformer decommissioning; emergency response;
                                                                          mobile incineration; and groundwater decontamina-
                                                                          tion.
                                                                          ERT
                                                                          696 Virginia Rd.
                                                                          Concord, MA 01742
                                                                                                            617/369-8910
Complete environmental consulting and technical/
engineering services—air, water, waste. Hazardous
waste management and consulting services: Super-
fund and RCRA consulting; waste  analysis  and
delisting assistance; audits; permitting; closure and
post-closure planning;  inactive waste disposal site
assessments/characterizations; remedial action pro-
grams; groundwater modeling and monitoring; risk
assessments; toxic air  pollutant measurement  and
analysis; emergency response systems.
                                                                                                                             1984 EXHIBITORS
                                                                                                                                                             615

-------
ESE, Inc.
P.O. Box ESE
Gainesville, FL 32604                904/332-3318
ESE, a full service multidisciplinary environmental
engineering firm, has performed work al more than
120 hazardous waste sites including  15 CERCLA
NCP sites. Capabilities include: remedial investiga-
tions;  feasibility studies; QA/QC plans; safety and
health planning and monitoring; community rela-
tions;   analytical  services;  and  expert   witness
testimony.
 E.G. Jordan Co.
 562 Congress Street
 Portland, ME 04112
                       207/775-5401
 Solid and hazardous  waste  management services
 provided  to industry and government agencies in-
 clude: geophysical  and geohydrological  investiga-
 tions; record  searches; chemical characterization;
 contamination risk  assessment;  identification  and
 evaluation of remedial action alternatives and im-
 plementation plans at hazardous waste sites. Hazar-
 dous waste TSD facilities are developed from initial
 planning stages, through site selection and investiga-
 tion, design,  permit application and construction
 management.
 Ecology and Environment, Inc.
 P.O. Box D
 Buffalo. NY 14225
                       716/632-4491
 Ecology and Environment is America's leading con-
 sultant for management of hazardous wastes and en-
 vironmental emergency  responses.  Complete Held
 investigation, monitoring,  cleanup,  analytical and
 engineering services are available for industrial and
 government clients throughout the United States
 and abroad.

 E.I. duPonl deNemours & Co.
 Tech. Lab., Chambers Works
 Deepwater, NJ 08023               609/540-3884
 Wastewater treatment service.
 Element Analysis Corp.
 16% Capital Circle SW
 Tallahassee, FL 32304               904/576-5115
 Element Analysis Corporation (EAQ is a commer-
 cial analytical laboratory featuring analysis by pro-
 ton induced  x-ray  emission (PIXE).  PIXE is a
 simultaneous, non-destructive technique for analyz-
 ing a sample for the elements from sodium through
 uranium. Sample matrices  may  vary  from solids
 (soils, powders, etc.) to liquids and aerosol fillers. In
 addition to  PIXE, EAC also offers a wide variety of
 other  analytical techniques, including atomic spec-
 troscopy. gaschromotography and wet laboratory
 techniques.
 Engineered Textile Product*, Inc.
 P.O. Box 7474                    205/476-8001
 Mobile, AL 36607                 800/222-8277

  E.T.P. is  a manufacturer/fabricator  of industrial
  textile products. The principal products arc pit and
  pond  liners,  floating  baffles, covers,  large  tar-
  paulins, etc.  Main  plant  offices  Mobile, AL.—
  branches in other cities.
 Engineering-Science
 57 Executive Park South, Suite 590
 Atlanta, GA 30329
                       404/325-0770
 Engineering-Science is a leading international en-
 vironmental engineering firm offering full services in
 solid waste management, hazardous waste remedial
 design  and implementation, permitting,  air pollu-
 tion  services,  waste  water  treatment,  hydro-
 geological and geophysical studies, field services and
 laboratory analyses. Through offices in major cities,
 Engineering-Science provides services to clients  in
 the government, military and private sectors.
                                           Evergreen Industries
                                           6390 Joyce Drive
                                           Golden, CO 80403
                                   800/525-8696
Portable  personnel  decontamination   units  for
hazardous  materials.  Manufactured  according  to
customer specifications. No decontamination unit is
too large or too small for Evergreen Industries to
service your decontamination needs. We can handle
everything  from asbestos to radioactive material
correctly.

Federal Emergency Management Agency
500 C Street,  S.W.
Washington, DC 20472               202/287-0300
The  exhibit   is  titled   "Integrated Emergency
Management  Systems." This exhibit  features elec-
tronic  displays of potential hazard  areas in  the
United States.
                                           Fram Industrial Filler Corporation
                                           Post Office Box 33210
                                           Tulsa, OK 74153                    918/252-9571
                                           Fram Industrial coalescing plate separator] will be
                                           exhibited. Fram Industrial manufacture! a complete
                                           range of gravity and pressurized separators. There
                                           are no filter cartridges or disposable elements used
                                           in  Fram  CPS  systems.  Many hazardous waste
                                           elements carried  by wastewater can be effectively
                                           removed in a Fram CPS Separator system. Fram ex-
                                           perience can help you win the wastewater battle.
                                           Fred S. James & Co., Inc.
                                           230 W. Monroe St., 19th Floor
                                           Chicago, IL 60606
                                  312/726-4080
                                           "Act 1, Scene II"—Fred S. James & Co.. a proven
                                           leader in the insurance brokerage business, is able to
                                           provide your company with a complete range of
                                           brokerage and insurance consultation services, in-
                                           cluding expert advise on environmental impairment
                                           liability insurance and  technical  assistance in the
                                           areas of hazardous  material  handling control and
                                           disposal, all designed to ensure that your company is
                                           in compliance with existing federal and state finan-
                                           cial responsibility laws  and that its assets  are ade-
                                           quately protected in the  event of  an unexpected
                                           release of pollutants or contaminants into the en-
                                           vironment.
                                           GAI Consultants, Inc.
                                           570 Beatty Rd.
                                           Monroeville. PA 15146
                                                                             412/856-6400
                                           GAI Consultants, Inc. and its subsidiaries provide
                                           engineering consulting services in the areas of solid
                                           and hazardous waste management, federal and state
                                           permitting assistance, and disposal site design ser-
                                           vices including remedial.investigations and feasibili-
                                           ty studies, geohydrologic investigations,  site selec-
                                           tion  and cost optimization  evaluations, and site
                                           operation and closure plans.
                                                       CCA/Technology Division
                                                       213 Burlington Road
                                                       Bedford, MA 01730
                                                                                          617/275-5444
CERCLA-related remedial engineering, field sam-
pling,  laboratory  analysis  and  groundwater
monitoring and modeling. Mobile hazardous waste
analytical  laboratories. Site  investigations and air
toxics monitoring. Complete  RCRA permit applica-
tion assistance. Incinerator trial burns.  Closure and
post-closure plans.  Registered engineers, geologists
and   industrial  hygienist.  AIHA  Certified
Laboratory.

GaUon Technical Services, Inc.
6601 Kirkville Rd.
E. Syracuse, NY 13057              315/432-0506
Galson Technical Services, Inc. was founded in  1970
to provide services  in i the field  of environmental
science. The Galson staff includes experienced en-
vironmental scientists,  engineers,   chemists,  in-
dustrial hygienists and meteorologists. Galson offers
services from our San Francisco and Syracuse of-
fices in the areas of: industrial hygiene, laboratory
analysis, source emission testing,  dispersion model-
 ing, meteorological studio, ambient air monitoring,
 environmental impact assessment  and hazardous
 waste management.
                                                                                                Geo-Con, Inc.
                                                                                                Post Office Box 17380
                                                                                                Pittsburgh. PA 15235
                                   412/244-8200
 Specialty construction service  company in  hazar-
 dous waste containment contracting, including pond
 liners, slurry walls, grouting and other techniques
 related to containment of fluids and underground
 seepage.

 Geonlc* Limited
 1745 Meyerside Drive. Unit #8
 Mississauga. Ontario
 L5T IC5 Canada                    416/676-9580
 Geonics Limited is the world's leading manufacturer
 of  electromagnetic geophysical equipment for the
 direct detection of ore bodies and geological map-
 ping. These same geophysical methods have proven
 to be invaluable for the detection of buried metallic
 drums and for mapping groundwater contaminant
 plumes. A wide range of ground conductivity meters
 can be seen al the Geonics Booth.

 Gronndwiler Decontamination System, lac.    103
 12 Industrial Park
 Waldwick, NJ 07463                 201/445-3141
 Patented  in situ  process for biodegrading  hydro-
carbons  and  halogenated  hydrocarbon  con-
taminants in  ground and  groundwater. The CDS
system  effectively decontaminates both soil and
groundwaler. making it more efficient, faster and
less expensive than other available  methods.  It is a
complete solution.
                                                    Goodie Limtag System*. Inc.
                                                    1340 E, Richey Road
                                                    Houston. TX 77073
                                                                                       713/443-8564
                                                    Gundle Lining Systems, Inc., a full service company
                                                    manufacturing,  installing and welding on-sile high
                                                    density polyethylene (HOPE) sheeting in thicknesses
                                                    ranging from 20 mil to 100 mil and widths of 22.5 ft.
                                                    These flexible membrane liners are used for applica-
                                                    tions pertaining  to environmental protection such as
                                                    irrigation dams, reservoirs, sanitary landfills, sewage
                                                    containment, earth canals and hazardous and non-
                                                    hazardous waste containments. Gundle has also an-
                                                    nounced their manufacturing of two new products.
                                                    Online and Hyperlastic.
                                                    HAZCO
                                                    1347 E. Fourth St.
                                                    Dayton. OH 45402
                                   513/222-1277
HAZCO is a total service supplier of all equipment.
training and program development needed for safe
handling of hazardous materials.  Services include:
safety/health  equipment sales; rental of decon and
equipment trailers, OVAs, HNU and SLBA. HAZ-
CO's Techservice is available at 1-800-332-0435 to
offer alternatives  for  specific safety/health con-
cerns.

tUM/HolznucDer, McLendon A
  Murrell, P.C.
125 Baylis Road, Suite 140
Melville, NY 11747                 516/752-9060
Environmental  analyses  including  industrial/
hazardous wastes, air, water, wastewater, sludge,
soils, leachate, dredge  spoils. Capabilities include:
bacteriology;  wet  and automated chemistry;  gas
chromatography; atomic   adsorption spec-
irophotometry and GC/MS. Consulting engineering
and  environmental  services including: impact
assessments;  NPDES  monitoring;  treatment/dis-
posal technology; RCRA  compliance;  regulatory
permit administration. USEPA  and  State(s) Ap-
proved Laboratory.
                                                                                               HNU Systems, Inc.
                                                                                               160 Charlemom St.
                                                                                               Newton, MA 02161
                                   617/964-6690
                                                                                               Model 301 gas chromatograph—a portable version
                                                                                               of the 301 with DC power/gas supply option pro-
616
1984 EXHIBITORS

-------
vines tield investigation capabilities. This G.C  can
be equipped with both  photoionization  detector
which  provides  field  confirmation  of  unknown
species. Model PI-101  hazardous waste  detector
provides  immediate indication of the presence of
potential toxic chemicals, even at sub ppm levels,
and gives a breakdown of existing compounds when
used with three available probes.

Harding Lawson Associates
7655 Redwood Blvd.
P.O. Box 578
Novato, CA 94948                   415/892-0821
Services:  hazardous  waste  site  investigations;
hydrogeological investigations; aquifer restoration;
leak detection; plume mapping; feasibility studies;
risk  assessment;  environmental  audits;  facility
design.  Disciplines:  hydrogeology; hydrology;
geology;  geochemistry; geophysics; environmental
engineering;  geotechnical  engineering;   chemical
engineering.

Hazardous Materials Control
  Research Institute
9300 Columbia Boulevard
Silver Spring, MD 20910             301/587-9390
HMCRI is a unique, public, nonprofit, membership
organization which promotes the establishment and
maintenance of a reasonable  balance between ex-
panding industrial productivity and an acceptable
environment. Our goals are met through a variety of
publications, conferences,  workshops, newsletters,
equipment  exhibitions  and   other  information
dissemination programs. We provide  members and
all other interested persons with a distinctive forum
in which they can exchange  information  and ex-
periences dealing  with hazardous materials.  Join
HMCRI today!!

Hazardous Waste Technology
  Services (HazTech)
3300 Marjan Drive
Atlanta, GA 30340                   404/451-9877
HazTech specializes in hazardous waste site cleanup
and emergency spill response.  Its offices in Atlanta
and Tampa are staffed with teams of professionals
in operation's  crews  that are highly  experience in
sampling, handling,  treatment and removal  of
hazardous materials. HazTech is currently  an EPA
Emergency Response Cleanup  Services (ERCS) con-
tractor.
 Health Evaluation Programs, Inc.
 808 Busse Highway
 Park Ridge, IL 60068
312/696-1824
 Health Evaluation Programs, Inc. provides nation-
 wide Mobile Health Testing Services, Health Status
 Profile Appraisals, a  Hearing Conservation  Pro-
 gram, and lab analysis for drugs of abuse, alcohol
 and marijuana. HEP is also introducing a new pro-
 gram this year called "Back in Health"—a back in-
 jury program.

 Hoyt Corporation
 251 Forge Road
 Westport, MA 02790                617/636-8811
 The Hoyt Odor-Miser is a prefabricated granulated
 activated carbon filter engineered  to remove low
 concentration odors, toxic substances, irritants, cor-
 rosive vapors and other gases  from vent systems.
 Odor-Misers  under  normal  circumstances  will
 eliminate obnoxious odors  from a wide range  of
 solvents and other organic and inorganic emissions.
 The Hoyt Still is a solvent purification and distilla-
 tion system engineered for simple solution to con-
 taminated solvent disposal problems. This unit has a
 stainless steel distillation tank, stainless steel  con-
 densor coil and heater coil.

ICF Incorporated
 1850 K Street, NW, Suite 950
Washington, DC 20006              202/862-1100
ICF is an employee owned consulting firm specializ-
ing in economic policy,  engineering and scientific
analyses for public and private clients. The 150 con-
                    sultants in environmental practice focus on  hazar-
                    dous substance control and waste management, pro-
                    viding risk  assessments, cost  analyses, regulatory
                    analyses, and  related services to our clients. Other
                    major environmental consulting analyses include
                    toxic substances, air and water pollution control.
                   ICOS Corporation of America
                   4 W. 58th St.
                   New York, NY 10019
                                   212/688-9216
                   Slurry walls,  slurry trenches,  drilling,  grouting,
                   bored piles, load bearing elements, tieback anchors,
                   sewer rehabilitation, shotcrete, Envirowall.
                   ISCO, Environmental Division
                   P.O. Box 82531
                   Lincoln, NE 68501
                                                      402/474-2233
                   Dedicated and portable well samplers, wastewater
                   samplers  and open channel flow meters will be ex-
                   hibited. The Model 2600 Well  Sampler, which has
                   pumping  rates to 2.2 gpm, is  designed to fit into
                   2-inch well casings. Well caps and quick disconnect
                   fittings are available for dedicated installations.
                   Industrial Training Systems Corp.
                   823 East Gate Drive
                   Mt. Laurel, NJ 08054
                                   609/234-2600
                   Industrial  Training  Systems   Corp.  designs,
                   develops, produces and markets mixed media train-
                   ing materials  in the area of environmental and oc-
                   cupational health and safety. We offer training pro-
                   grams  on RCRA,  TSCA, Compliance Awareness
                   and a  seven-part  series (live video) on hazardous
                   spill management. Also programs  on health hazard
                   communication employee training.
                   In-Situ, Inc.
                   209 Grand Avenue
                   Laramie, WY 82070
                                   307/742-8213
                   Computer-automated groundwater level monitoring
                   instrumentation.  Multidisciplinary  services:
                   metallurgical and analytical laboratories; hydrologic
                   and geotechnical consulting; feasibility studies and
                   process  design; computer timesharing; hydrologic
                   and  energy related  software  for  the IBM  PC;
                   graphic software; oil and gas and mining software.

                   International Engineering Company
                     Monison-Knudsen Co.
                   180 Howard Street
                                                     San Francisco, CA 94105
                                                                                        415/442-7300
International   Engineering  Company/Morrison-
Knudsen  Company provides  site  investigations,
remedial  designs, construction management  and
remedial actions. Site investigations include: ground
and  surface   water  monitoring;  definition  of
hydrologic regime; site geology and  contaminant
transport. Remedial design includes: slurry trench
design;  encapsulation; site drainage improvements;
cost estimates; specifications; drawings and feasibili-
ty studies.

International Marketplace/
   Hospitality Booth
A distinctive  area for  conference  attendees, ex-
hibitors, guests, special dignitaries and international
representatives  to  convene during  the  Conference
and Exhibition. HMCRI is extremely pleased  and
proud to  have this important and timely interna-
tional portion  of  the 5th Superfund  Convention,
and everyone in attendance is  encouraged to visit
this area to become more acquainted  with the ac-
tivities of our international counterparts.

JRB Associates
A Company of SAIC
8400 Westpark  Drive
McLean, VA 22102                703/821-4886
JRB,  a  Company of Science Applications Interna-
tional Corporation, is pleased to demonstrate its ex-
perience and expertise in the management of hazar-
dous  wastes  including  site investigation, RCRA
"Part B" Permit applications, chemical  industry
studies,  technology evaluation for  hazardous spill
                                                       cleanup and site remediation. Additional services in-
                                                       clude  environmental  audits,  expert  witness
                                                       testimony and regulatory development.
                                                       James T. Warring Sons, Inc.
                                                       4545 "S" Street
                                                       Capital Heights, MD 20743
                                                                                                             301/322-5400
 All types and sizes of containers—new & recondi-
 tioned—fiber, steel, plastic. Our  hazardous waste
 containers are DOT approved and range in size from
 5 to 83 gallons. We accept orders from one to truck
 loads  and we ship  anywhere. You order  a  con-
 tainer—we don't have it—it's special—we will get it
 for you. No order is too small for James T. Warring
 Sons,  Inc. Let us help you contain your hazardous
 waste. We can it!
                                                       J.J. Keller & Associates, Inc.
                                                       145 West Wisconsin Avenue
                                                       Neenah, WI54956
                                   414/722-2848
 Technical publisher and  regulatory services com-
 pany specializing in consulting and safety services,
 transportation  and   hazardous  material/waste
 publications, forms and supplies. Our product line
 includes:  guides,  manuals,   training   materials,
 placards,  labels, spill control  devices, wall charts,
 hazardous waste manifests, and bills of lading. We
 also offer a full line of products to solve most any
 transportation or  hazardous material/waste prob-
 lem. We can service  your needs.  Phone: national
 800/558-5011;  Wisconsin  800/242-6469;  local
 414/722-2848.
K.W. Brown and Associates, Inc.
6A Graham Rd.
College Station, TX 77840           409/693-8716
K.W. Brown and Associates,  Inc. (KWB&A) has a
multidisciplinary staff of professionals who can pro-
vide the state-of-the-art expertise needed to develop
solutions to your environmental management and
pollution needs. KWB&A specializes in soil related
aspects of the storage, treatment and  disposal  of
both hazardous and nonhazardous wastes, cleanup
assessment of salt and chemical spills, compatibility
testing of clay liner-waste combinations, reclama-
tion of disturbed lands and interpretation  of soil
analyses.  Through   research   and  consulting,
KWB&A has developed an  in-depth understanding
of the movement and degradation of chemical com-
pounds and plant nutrients in  the soil environment,
which is an essential part of the solution to many en-
vironmental   problems.  KWB&A has  provided
assistance to numerous  clients, including petroleum
refineries, chemical plants, waste disposal and min-
ing companies,  manufacturing facilities, other en-
vironmental  consultants,  law firms, public interest
groups, individuals,  and federal, state and local
government agencies.
                                                                         LOP AT Enterprises, Inc.
                                                                         640 Mattison Avenue, Suite 200
                                                                         Asbury Park, NJ 07712
                                                                                                            201/776-6710
                                                                         Lopat's K-20  (patent pending) is a non-volatile,
                                                                         non-corrosive, inorganic colorless and odorless for-
                                                                         mulation.  It   has   been  proven  effective  on
                                                                         chloradane and PCBs commercially. Laboratory
                                                                         tests prove the product to be totally effective as an
                                                                         incapsultant of asbestos. K-20 was used commercial-
                                                                         ly against PCBs by IT Corporation on five different
                                                                         occasions and in one instance reduced levels from 60
                                                                         ppm to 0.1 ppm. It has been used effectively against
                                                                         chloradane reducing levels  from  as high  as 1037
                                                                         cc/sq. ft to non-detectible as per a test done for the
                                                                         Dept. of  Agriculture,  Pesticide Board,  State of
                                                                         Massachusetts. Samples of Lead contaminated soil
                                                                         received from Dallas, TX with readings of 200 ppm
                                                                         were reduced to 0.1 ppm. K-20 is a penatrant, not a
                                                                         surface sealant, reaching depths of anywhere from
                                                                         .75 in.  to  2.0 inches in concrete and cinder block.
                                                                         K-20 appears in a forthcoming EPA paper entitled
                                                                         "Practical Methods for Decontaminating Buildings
                                                                         and Structures at EPA Superfund  Sites."
                                                                                                                            1984 EXHIBITORS
                                                                                                                         617

-------
Law Engineering Testing Co./
  Law Environmental Services
1140 Hammond Drive, E-5150
Atlanta, GA 30328
                                  404/396-8000
Law Engineering Testing Company is an interna-
tionally established environmental consulting firm.
Law's  team  of environmental professionals  is sup-
ported by specialized equipment, laboratories and
computer  capabilities to provide the full range  of
services necessary to locate, design, permit and
operate commercial and industrial facilities.

MAC Corporation/Saturn Shredder
  Division
201 East Shady Grove Road
Grand Prairie, TX 75050             214/790-7800
MAC  Corporation/Saturn   Shredder   Divi-
sion—Shear type shredders for shredding of hazar-
dous waste,  municipal waste, inplant waste, wood,
tires, and  various other materials for size reduction.
Marine Pollution Control
8631 West Jefferson Avenue
Detroit, MI 48209
                                   313/849-2333
 Marine Pollution  Control was  one  of  the  first
 cleanup companies in the United States.  We have
 developed into a mobile, rapid  response oriented
 company, capable of responding to diverse condi-
 tions.  We are able to handle oil and  chemical in-
 cidents in  land or  waier environment.  We  now
 possess high  capacity  pumping  for emergency
 response conditions—3000 GPM for light products
 and capable of pumping coal tar.

 Mateson Chemical Corporation
 1025 East Montgomery Avenue
 Philadelphia, PA 19125             215/423-3200
 Specialists in clean air and surface restoration after
 hazardous materials "accidents" (spills, contamina-
 tion, etc.)-  Manufacturers of ecologically safe and
 test-proven chemicals for decontamination, detox-
 ification, encapsulation, neutralization, adsorption,
 deodorization, disinfecting and other odor-sorbing
 products, and we provide "hi-tech" expertise and
 knowledge   in  all  these  areas.  We  are  also
 distributors  for air,  particle and gas analyzers,
 HEPA Air Cleaners,  HEPA  Vacuums,  mobile
 decontamination   stations,  hi-pressure  washers,
 sprayers, foggers, etc.
 Med-Tox Associates, Inc.
 1401  Warner Avenue, Suite A
 Tustin, CA 92680
                      714/669-0620
 Med-Tox Associates,  Inc.  offers services in In-
 dustrial  Hygiene,  Toxicology  and  Occupational
 Medicine. Health and safety plans, generic and site
 specific  are  developed.   Toxicological  risk
 assessments and hazardous  materials management
 are provided, along with training programs. Medical
 services include the development  and  implementa-
 tion of medical standards and medical surveillance
 programs.

 Metcalf & Eddy, Inc.
 50 Staniford Street
 Boston, MA 02114                 617/367-4004
 Metcalf  & Eddy,  a  multidisciplinary  team  of
 engineers,  scientists and  health specialists  are pro-
 viding services to governmental and private entities,
 responsible for hazardous waste management. Met-
 calf & Eddy has become a leader in hazardous waste
 management, including hazardous waste site clean-
 up and waste treatment and disposition projects and
 technologies. Metcalf & Eddy  has conducted site
 remedial investigations, endangerment assessments
 and feasibility studies and provided design and con-
 struction  management  services  at  numerous in-
 dustrial hazardous waste landfill sites and  national
 priority listed sites (NPL), throughout the United
 States  and  abroad.   Our  hazardous   waste
 specialists—toxicologists,  chemists,  chemical and
 environmental   engineers,  soil  scientists,
 hydrogeologists, surface water, air and groundwater
 modelers, and  health and safety  experts—all give
 Metcalf &  Eddy the ability  to conduct  hazardous
                                         waste site studies, and  to  design and implement
                                         remedial  clean-up activities which would mitigate
                                         any waste site threats to public health, welfare and
                                         the environment. Our specialty includes appropriate
                                         technology evaluations, health and safety planning,
                                         remedial  design,  construction  management, en-
                                         vironmental   impact   assessments,   endangerment
                                         assessment, feasibility studies, remedial  investiga-
                                         tions and source reduction engineering. For more in-
                                         formation   contact:  Donald   P.   Brown,
                                         617/367-4395.

                                         NUS Corporation
                                         910 Klopper Road
                                         Gailhersburg, MD 20878             301/258-1299
                                         The Site Remediation and Construction Division of
                                         NUS combines the scientific and engineering exper-
                                         tise of NUS  with the  construction and  project
                                         management capabilities of  Brown & Root to pro-
                                         vide  a single source  for  hazardous  waste site
                                         remediation. Services range from remedial investiga-
                                         tion/feasibility studies through detailed engineering
                                         and construction management. NUS and Brown  &
                                         Root are well aware of potential liabilities associated
                                         with hazardous waste and are dedicated to achieving
                                         a sound technical solution to your hazardous waste
                                         problem  at a  reasonable cost. Our  single respon-
                                         sibility approach  assures  client  confidentiality
                                         through project completion. Feasibility studies are
                                         done with ultimate site remediation cost in mind. As
                                         a result,  site  remediation  can  be  accomplished
                                         thoroughly and quickly.

                                         National Audiovisual Center
                                         National  Archives
                                         Washington, DC 20409              202/763-1850
                                         The  National Audiovisual  Center—the  central
                                         source for Federal audiovisual programs—will be
                                         showing  the  in-depth  self-teaching   program,
                                         "Transportation of Hazardous Materials." This II
                                         unit course will: meet  the training requirements of
                                         CRF Titles 14 & 49; tell who must comply with the
                                         federal regulations;  give consequences of non-
                                         compliance; identify HM; show proper packaging,
                                         transporting  and incident reporting  requirements.
                                         Stop by booth 1804 and examine the complete pro-
                                         gram  as  well  as  many  other  related  programs on
                                         display.
                                                      National Library of Medicine
                                                      8600 Rockvillc Pike
                                                      Bethesda, MD 20209
                                                                           301/496-1131
The NLM Chemical and Toxicological Files are an
online, interactive  retrieval service in toxicology.
They  include CHEMLINE (Chemical  Dictionary
Online),   TOXLINE  (Toxicology  Information
Online),  RTECS (Registry of Toxic  Effects of
Chemical Substances), and TDB (Toxicology Data
Bank). These files are a pan of the Library's
Medical Literature Analysis and Retrieval System
(MEDLARS).
                                         National Seal Company
                                         600 North First Bank Drive
                                         Palatine, IL 60067
                                                                            312/991-6929
                                         National Seal Company is a leader in the manufac-
                                         ture and  installation  of geomembranes featuring
                                         turnkey projects. With increased emphasis on quali-
                                         ty control, especially  in hazardous waste contain-
                                         ment, NSC has excelled by providing state-of-the-
                                         art seaming and installation technology of its mem-
                                         branes.  NSC's booth  will display seam samples of
                                         various  geomembranes, highlighting  HDPE  and
                                         XR-5 along with brochures describing our complete
                                         line of liners and turnkey capabilities.
                                         National Spill Control School
                                         6300 Ocean Drive
                                         Corpus Christ!, TX 78412
                                   512/991-8692
                                         The National Spill Control School provides hazar-
                                         dous materials/hazardous waste training at Corpus
                                         Christi and on-site. Courses include: Spill Preven-
                                         tion and Control  training  for managers  and Site
                                         Safety Training, a  hands-on course, for responders
                                         and handlers. On-site courses are developed for the
                                         specific  needs of the organization.
                                                     O.H. Materials Co.
                                                     P.O. Box 551
                                                     Findlay, OH 45839                  419/423-3526
                                                     Hazardous  materials/substances containment and
                                                     cleanup; environmental  restoration; planned and
                                                     emergency remedial actions.

                                                     ONTECH, Inc.
                                                     Post Office Box 171224
                                                     Arlington, TX 76003                817/861-5307
                                                     ONTECH,  Inc. provides industry  with  computer
                                                     software  and  hardware tools to improve en-
                                                     vironmental  management  and  pollution  control
                                                     operations. ONTECH markets a computer program
                                                     called  IRIS,  the Industrial Resources Information
                                                     System that  will:  I.  Prepare and  print shipping
                                                     manifests; 2. Maintain permanent storage of ship-
                                                     ping and waste disposal  information; 3. Catalog
                                                     transporter and disposal sites; 4. Provide rapid ac-
                                                     cess to safety and emergency information; 5.  Print
                                                     monthly and  annual  reports of  shipping and
                                                     disposal operations; and 6. Provides a quantitative
                                                     analysis of waste management operations and costs.
                                                    Oil Recovery Systems, Inc./
                                                      Groundwaler Technology
                                                    299 Second Avenue
                                                    Needham, MA 02194                617/449-5222
                                                    Equipment and  full  services to solve groundwater
                                                    contamination  problems:  Scavenger*  recovery
                                                    systems, water  table  depression  pumps,  water
                                                    purification and vapor recovery systems, interface
                                                    probes and monitoring equipment. Services include:
                                                    cost-effective recovery design; assessment; emergen-
                                                    cy response, and risk management. Offices nation-
                                                    wide.
                                                    Oxford Linen Inc.
                                                    Post Office Box 507
                                                    Williamsville. NY 14221
                                                                                       716/688-1321
                                                    With increasing costs, many are electing to install
                                                    polyethylene liners with in-house staff. Having this
                                                    in mind, Oxford Liners, Inc., has available technical
                                                    and equipment  assistance for expert quality liner
                                                    systems. Our technical  staff and a strong compli-
                                                    ment of installation  equipment permits rapid  in-
                                                    stallation by anyone,  anywhere in  North America.
                                                    Call us for a representative in  your area. See our
                                                    welding technology continuously demonstrated at
                                                    Booth 707.
                                                                                              PB KBBInc.
                                                                                              P.O. Box  19672
                                                                                              Houston, TX 77224
                                                                                                                                713/496-5590
Subsurface  engineers  specializing  in  the
underground disposal of hazardous waste—especial-
ly  in bedded  or domal salt formations. Company
capabilities include permitting, design, construction
management and facility operation.

Packaging Research and Design Corporation
12717 Pecos Ave.
Greenwell Springs, LA 70739         504/261-1474
Packaging  Research  and  Design  Corporation
specializes in  custom designing and manufacturing
disposal  plastic  bags for transporting hazardous
wastes. Our patented bag liners are offered in many
different sizes for roll-offs, dump trailers, vans and
luggers.


Penberthy Eleclromell International, Inc.
631 South 96th St.
Seattle, WA 98108                  206/762-4244
The Penberthy Pyro-ConverterTM is a tunnel in-
cinerator having a pool of molten glass covering the
bottom. The  molten  glass, heated electrically, con-
stantly maintains required destruction temperature
no matter what is being burned  (oil,  solvents, car-
bon let, PCB, water, mud, dirt).  Inorganic material
dissolves  in  the  glass and  becomes a delisted
material.
 618
1984 EXHIBITORS

-------
Perma-PIpe
7720 Lehigh Avenue
Niles.IL 60648
                                   312/966-2150
Perma-Pipe  offers  Double-Pipe/PermAlert  II
systems designed exclusively for the safe transporta-
tion of hazardous fluids. Featuring Secondary Con-
tainment and Leak Detection, they will: contain
hazardous fluid leakage, preventing it from entering
the environment; protect service pipes  from cor-
rosive external conditions, detect leaks and provide
audio/visual  alarms;  locate  leaks immediately
displaying and recording location and time of occur-
rence.  Double-Pipe is one sure way to protect the
environment, public health and your bottom line.

Photovac Incorporated
Unit 2, 134 Doncaster Avenue
Thomhill, Ontario
L3T1L3 Canada                    416/881-8225
Photovac's Model 10A10 Portable GC has become
the tool of choice in many hazardous waste site in-
vestigations. Allows rapid screening of air, ground-
water and soil for trace volatiles; these include vir-
tually all chlorinated solvents. Photovac's new TIP
product  provides  a  very  portable,   low cost,
photoionization scoping tool  for preliminary in-
vestigations.

Pollution Abatement Consultants &
   Services
Div. of Wheaton Ind.
1301 N.  lOih St.
 Millville.NJ 08332
                                   609/825-1400
 PACS will have on display hazardous material ship-
 ping containers and packages; samplers for waste
 streams, ponds, lagoons,  sludges & soils,  drums,
 etc.; safety coated glass sample containers,  general
 solvent analysis/solvent recovery units, lab  aids to
 meet EPA & ASTM standards, and portable in-
 struments for field use.

 Poly-America, Inc.
 2000 West Marshall Drive
 Grand Prairie, TX 75051             817/640-0640
                                   800/527-3322
 Poly-Flex Geomembranes—many  advantages in-
 clude:  resistance to most chemicals; high  tensile
 strength;  stress  crack  resistance;  cost effective;
 widest seamless sheet; ease of installation and easily
 sealed. Poly-Flex's combination of properties makes
 it the best liner choice for most applications.
 Princeton Aqua Science
 789 Jersey Avenue, FOB 151
 New Brunswick, NJ 08902
201/846-8800
 Environmental consulting and testing. Complete en-
 vironmental laboratory services, plus expert con-
 sulting  by engineers and  scientists on  industrial,
 hazardous waste,  aquatic,  terrestrial  and  en-
 vironmental sciepce projects.
 Princeton Testing Laboratories
 Post Office Box 3108
 Princeton, NJ 08540
609/452-9050
 Princeton  Testing  Laboratory  is a  leading  en-
 vironmental and material analysis laboratory with
 five different  sections. There  is a Spectrographic
 Laboratory, a Water, Wastewater, & Microbiology
 Laboratory, an Inorganic Laboratory,  an Organic
 Laboratory, and an Industrial Hygiene & Air Pollu-
 tion Laboratory. We are NJDEP certified  and do
 analysis  on  drinking  water,  potable  water,
 wastewater, sludge, sewage, soil, toxic and hazar-
 dous waste  for  inorganic impurities  and  con-
 taminants. Also analyzed are environmental samples
 for trace, toxic or hazardous organic pollutants by
 GC, HPLC and GC/MS. Commercial products are
 analyzed for low level concentrations of impurities
 as well as major constituents, moisture content and
 contaminants. Our Industrial Hygiene section does
 ambient air testing and surveys, exhaust gas testing
 and certification, workplace  environment testing
 for OSHA regulated contaminants and air quality
 testing of home and office.
                   QED Environmental Systems, Inc.
                   1254 N. Main St.
                   Ann Arbor, MI 48107               313/995-2547
                   Well Wizard groundwater sampling system. Speeds
                   and improves the collection of groundwater samples
                   from monitoring wells as small as 2-inch diameter
                   and as  deep as 230 ft. Each well is equipped with a
                   down well pump which is operated by an easy-to-use
                   portable controller. The Well Wizard is the only
                   dedicated well sampling system  and is specifically
                   engineered to ensure sample integrity and reduce
                   field labor.
                   REACT
                   P.O. Box 27310
                   St. Louis, MO 63141
                                                     -314-569-0991
                   REACT environmental crisis  engineers offers na-
                   tionwide 24-hour emergency response for hazardous
                   material accidents; environmental crisis engineering
                   for the containment and control of hazardous waste
                   sites;  and  consulting engineering,  including  in-
                   dustrial hygiene, RCRA compliance audits, ground-
                   water  contamination  studies,  SPCC  plan prepara-
                   tion and chemical hazard analysis and evaluation of
                   potential health effects.
                   Radian Corporation
                   P.O. Box 9948
                   Austin, TX 78766
                                                     512/454-4797
                   Radian Corporation  offers turn-key solutions to
                   hazardous waste management problems. A staff of
                   800 +,in strategically located offices throughout the
                   country, provides  expertise  in:  waste characteriza-
                   tion, site  investigations,  sampling  and  analysis,
                   remedial action feasibility, incinerator testing,  per-
                   mitting support, emergency  response planning, and
                   design of remedial action and waste control systems.
                   Recra Research, Inc.
                   4248 Ridge Lea Road
                   Amherst, NY 14226
                                                     716/838-6200
Recra Research, Inc. is a chemical, environmental
and engineering consulting firm with over 160 per-
sonnel in Buffalo, New York and Houston, Texas.
Three  laboratories  (environmental,  waste,
geotechnical) support the consulting groups. En-
vironmental compliance  with  state and  federal
regulations and cost  effective waste treatment and
management systems  are primary services offered to
industry.

Resource Analysts, Inc.
Post Office Box 4778
Hampton, NH 03842                603/926-7777
Resource  Analysts,   Inc.  is   an   environmental
chemistry service organization. Specialties include
organic, metals, and  other inorganic wet chemistry
analyses of air, water and solid  media in accordance
with environmental testing protocol. Consultation is
provided for scoping  work and interpreting results.
Field sampling support service  is  also available.
Laboratory is  certified under  EPA programs and
participates in site investigations, site remediation,
industrial operation  and  government enforcement
monitoring.
                   Resource Technology Services, Inc.
                   6 Berkeley Rd.
                   Devon, PA 19333
                                   215/687-4592
                   Resource Technology Services, Inc. is a hazardous
                   waste management  organization involved in  the
                   hands-on removal and disposal of chemical wastes.
                   RTS offers a complete range of services including
                   transportation, storage, consulting, personnel train-
                   ing, remedial action  and emergency response.
                   R.E. Wright Associates, Inc.
                   3240 Schoolhouse Road
                   Middletown, PA 17057
                                   717/944-5501
                   R.E. Wright Associates, Inc. is an applied ground-
                   water consulting firm employing professional scien-
                   tists  in the  field  of hydrogeology,  geochemistry,
                   engineering geology, soil science, aquatic  biology
                   and geophysics, and providing services to industries
                                                      involved in the manufacture, use and disposal  of
                                                      hazardous  materials and to State and Federal en-
                                                      vironmental agencies.
                                                      Rexnord
                                                      45 Great Valley Parkway
                                                      Malvern, PA 19355
                                   215/647-7200
                                                      Rexnord Electronic  product division  will exhibit
                                                      self-contained breathing apparatus for the FireEN-
                                                      TRY, ChemENTRY and HazMatENTRY applica-
                                                      tions. The BioPak 60 offers full one hour duration,
                                                      positive pressure  and lightweight—24 pounds. En-
                                                      viroEnergy  Technology  Center of  Rexnord has
                                                      special  experience  and  services  in  analytical
                                                      laboratory  bench tests and pilot physical chemical
                                                      and  biological systems and  operations  including
                                                      mobile vans. Full size, scaleup and implementation
                                                      including Spill Prevention Control  and Contain-
                                                      ment.

                                                      Roy F. Weston, Inc.
                                                        Designers/Consultants
                                                      Weston Way
                                                      West Chester, PA 19380             215/692-3030
                                                                                              ext. 257
                                                      Roy F. Weston, Inc.  is a leader in hazardous waste
                                                      and  remedial action  engineering providing com-
                                                      prehensive  environmental  management services in
                                                      the areas of: risk  assessments, compliance reviews/
                                                      audits,  permitting,   site  upgrading/mitigation/
                                                      cleanup, emergency response, air quality monitoring
                                                      and real-time environmental monitoring.
                                                      SCS Engineers
                                                      4014 Long Beach Blvd.
                                                      Long Beach, CA 90807-2687
                                  213/426-9544
SCS Engineers  specializes in  solid and hazardous
waste management issues. Services offered  by the
firm include:  selection  and  design of  remedial
measures; preparation of closure plans; preparation
of spill  response plans; and permitting assistance.
The firm is experienced in toxic air emission control
and treatment. Founded in 1970, SCS maintains of-
fices in  Long Beach, CA;  Reston,  VA; Bellevue,
WA; and Covington, KY.
                                                                        SIJAL Inc.
                                                                        205 Roesch Avenue
                                                                        Oreland, PA 19075
                                   215/572-0216
Protective clothing including Chemtex coveralls and
three-piece suits specially formulated for chemical
resistance.

SKC, Inc.
395 Valley View Road
Eighty Four, PA 15330              412/941-9701
SKC manufactures instruments for chemical hazard
detection and air sampling. Complete line of air
sampling pumps and tubes; calibration equipment;
asbestos  test  kit and collecting medium  and ac-
cessories,  i.e.,  filters,  impingers,  sample  bags.
Passive  dosimeters include: organic vapors, liquid
and specific badges for mercury and phosgene. New
products  include:   portable   colorimetric   tape
monitors for phosgene, hydrazines and isocyanates.

SMC Martin Inc.
900 W.  Valley Forge Rd.
P.O. Box 859
Valley Forge, PA 19482              215/265-2700
SMC Martin Inc., over the past two decades, has
imaginatively  combined consulting services in the
fields  of geology,  hydrogeology,  geochemistry,
soils, civil engineering, planning and surveying. Our
professionals regularly provide expertise in ground-
water protection and remediation,  water resources
development, mining, foundations and environmen-
tal assessment in a cost effective manner.
                                                      SRW Associates, Inc.
                                                      2793 Noblestown Road
                                                      Pittsburgh, PA 15205
                                   412/921-0321
                                                      SRW Associates, Inc. is a civil, geotechnical and en-
                                                      vironmental engineering firm specializing  in waste
                                                                                                                            1984 EXHIBITORS
                                                                                                                         619

-------
management and waste engineering tor industry and
government.  Services  include: design;  permitting;
groundwater  monitoring;  planning;  remedial in-
vestigation;  feasibility studies;  site  closure; soil
laboratory testing; liner compatibility testing, and
Part B applications.

Seaman Corporation
102 N. Washington St.
Millersburg, OH 44654              216/674-0040
Seaman  Corporation is  a manufacturer  of high
quality membrane lining products; products which
are used for hazardous waste containment including
landfill and liquid containment.

Senlex Sensing Technology, Inc.
553 Broad Avenue
Ridgefield, NJ 07657                201/945-3694
Services:  Identification and monitoring of known
and unknown hazardous vapors at various sites
through G.C.M.S. analysis and  computerized por-
table gas chromatograph.  Devices:  Computerized
portable  gas   chromalograph;  multi-point
chromatograph system.

Sevenson Construction Corporation
   Hazardous Waste Division
2749 Lockport Road
 Niagara Falls, NY 14302             716/284-043!
Sevenson Construction Corporation operates a divi-
 sion specializing in a  wide variety of remedial con-
struction techniques at hazardous waste sites. Ser-
vices performed include the removal,  treatment and
disposal of hazardous materials, earthwork, con-
 crete work, secured on-site and off-site containment
 facilities, lagoon solidification, slurry  walls, leachate
 and waste water treatment facilities
 Shepherd Engineering Testing Co., Inc.
 2720 South Classen
 Norman, OK 73071                405/364-2900
 Shepherd Engineering Testing Co., Inc. is a full ser-
 vice  geotechnical  engineering  and  construction
 materials testing company. The firm specializes in
 geotechnical drilling, monitor wells, piezometer in-
 stallation, and mineral exploration drilling activities
 with large emphasis on hazardous materials sites.

 Shirco, Inc.
 1195 Empire Central
 Dallas, TX 75247                   214/630-7511
 Shirco, Inc. incineration systems featuring the use of
 InfraRed heating and conveyor belt  transport of
 waste  materials  through an  efficiently insulated
 modularly constructed waste disposal system. Since
 no fossil fuel is  required, the reduced gas flow is
 economically treated to meet requisite emission stan-
 dards. Systems are excellent for intermittent opera-
 tion and have transportable capability. Shirco Por-
 table Pilot Test  Unit available for on-site testing at
 your facility.
Slurry Systems/American
  Foundation, Inc.
Post Office Box 360
Morrisville, NC 27560
                                  919/467-7896
                                  215/678-7176
Vibrated Beam Slurry Cut-Off Wall Systems—for
chemical/hazardous waste  and  leachate control,
water conservation and construction  dewatering.
Our  specially formulated  slurries impede  water
pollutants  to the specified  "K-factor"  of im-
permeability. We Cut Off Problems!

SolldTek Systems, Inc.
5371 Cook Rd.
P.O. Box 888
Morrow, GA 30260                404/361-6181
Full spectrum of hazardous waste handling, treat-
ment, and disposal services. Custom chemical pro-
ducts and services  for solidification, with chemical
fixation, for detoxification  and  delisting.  Mobile
services for  processing, remediation and emergen-
                                                     cies. Regional TSDF service. Installed systems for
                                                     generator locations and POTWs. Advanced secure
                                                     landfill.
                                                      Stablex-Reuller, Inc.
                                                      28 Springdale Rd., Bldg. 21
                                                      Cherry Hill, NJ 08034               609/751-1122
                                                      Stablex-Reutler is a  fully integrated environmental
                                                      analytical laboratory and consulting firm specializ-
                                                      ing in full priority pollutants and RCRA analyses.
                                                      S-R has a mobile  laboratory and  extensive field
                                                      sampling capabilities for hazardous site evaluations.
                                                      S-R is fully accredited  by USEPA/NJDEP  and
                                                      A1HA.   Our  instrumentation   includes:  4
                                                      GC/MS/DS, 6 GC, 3 AA,  I 1C AP, I 1C.
                                                      Stauffer Chemlcil Company
                                                      Nyala Farms Road
                                                      Westport, CT 06484
                                   203/222-3000
                                                      Stauffcr utilizes hazardous wastes  for their energy
                                                      value as NON-CONVENTIONAL FUELTM in their
                                                      Sulfuric Acid Regeneration plants.
                                                     TECHNOS, Inc.
                                                     3333 NW 21st Street
                                                     Miami, FL 33142                   305/634-4507
                                                     TECHNOS is a consulting firm in the applied earth
                                                     sciences, and specializes in subsurface investigations
                                                     for:  hazardous waste site investigation; mapping
                                                     landfill leachate;  permitting and monitor  plans;
                                                     post-closure long-term monitoring; geotechnical in-
                                                     vestigations; subsidence and sinkhole evaluation;
                                                     Karst hydrogeology; search and location of  buried
                                                     containers.

                                                     TRC Environmental Consultants, Inc.
                                                     Advanced Analytics, Inc.
                                                     800 Connecticut Boulevard
                                                     East Hartford, CT 06108
                                                                                       203/289-8631
                                                     TRC  Environmental Consultants specializes in in-
                                                     novative solutions to air, water and hazardous waste
                                                     pollution problems.  Hazardous waste services in-
                                                     clude determination  of contaminant  migration,
                                                     design of  remedial actions,  waste  management
                                                     engineering, site audits, and real-time monitoring of
                                                     toxic  pollutants using TAG A* , a mobile MS/MS
                                                     system owned by TRC Advanced Analytics.
Training & Information Services, Inc.
P.O. Box  4769
Silver Spring, MD 20904-0769        301/236-0409
T1SI  provides course  development,  training and
consultation in the areas of hazardous materials leak
and spill control, disposal, medical problems, oc-
cupational health, emergency  medical services and
fire service response, organization and education.

Trla
(Elson T.  Killam Associates, Inc.)
P.O.  Box 1008
27 Blceker Street
Millburn,  NJ 07041                 201/379-3400
Tria is a  technical association of Elson T.  Killam
Associates,  Inc.,  [ cggetlc, Brashears  &  Graham,
and   GEOMET   Technologies.  The  Tria  group
specializes  in  comprehensive  hazardous   waste
management   including  site, inspection,
hydrogeology, health  and  safety,  site  engineering
and remedial design.


U.S. Army Corps of Engineers
P.O. Box  103, Downtown  Station
Omaha, NE68I01                  402/221-7317
The U.S. Army Corps of Engineers and the USEPA
have joined  forces to clean up  Federal Lead hazar-
dous waste sites under the Superfund program. The
booth will be manned by Corps' personnel to assist
architect-engineer  firms and construction contrac-
tors lake advantage of work available to them under
Superfund through the  Corps of Engineers.
 U.S. Army Environmental Hygiene
   Agency
 Aberdeen Proving Ground
 Aberdeen, MD 21010                301/671-3651
 U.S. Army Environmental Hygiene Agency, Waste
 Disposal Engineering Division—Army and Depart-
 ment of Defense worldwide support on the manage-
 ment  and disposal of hazardous and solid  wastes,
 emergency spill response, soil analysis and ground-
 water monitoring.

 U.S. Environmental Protection Agency
 Raritan Depot
 Edison, NJ 08837
                                   201/321-6677
 The USEPA has  been actively involved at all levels
 of hazardous waste management with the investiga-
 tion, categorization, response activities and remedial
 actions  at hundreds  of  uncontrolled  hazardous
 waste sites. This  display exhibits  products  for the
 protection of response  personnel, and provides in-
 formation on current Superfund activities, the latest
 response equipment for site  investigation, research
 and development  for long-term remedial action and
 USEPA's mobile  treatment systems.
                                                      U.S. Geological Survey
                                                      790 National Center
                                                      Reston, VA 22092
                                                                                        703/860-6162
                                                     Panels depicting research and products of the U.S.
                                                     Geological Survey dealing with earth sciences.


                                                     Union Piclfk System
                                                     1416 Dodge Street
                                                     Omaha, NE 68179                   402/271-4715
                                                     Union Pacific System offers a rail alternative to
                                                     over-the-road handling of hazardous materials and
                                                     wastes. Trained  personnel  in hazardous materials
                                                     and waste handling, well defined incident response
                                                     procedures, economies in transport costs, and an ex-
                                                     cellent safety record make this a very attractive op-
                                                     tion to waste haulers  and disposers. Union Pacific
                                                     will assess your transportation needs and prepare a
                                                     special logistics and pricing  package.
\ elsicol Chemical Corporation
2603 Corporate Ave., Suite 100
Memphis, TN 38132                 901/345-1788
Velsicol Chemical Corporation is a medium-sized
chemical  company  with  2,000  employees doing
business on a worldwide basis. Velsicol's manufac-
turing plants, located in the United States, Mexico
and  Brazil,  manufacture  herbicides,  pesticides,
rodenticides,   intermediate  chemicals  and
plasticizers. A strong and growing company, one
key to Velsicol's business success is a corporate
philosophy that includes one of the strongest com-
mitments 10 environmental integrity the chemical in-
dustry has ever  witnessed.

WAPORA, Inc.
6900 Wisconsin Avenue
Chevy Chase, MD 20815             301 /652-9520
WAPORA, Inc. is  a leading engineering  and en-
vironmental consulting  firm that has served  industry
and government for the past 15 years. The  firm of-
fers a wide range of experience in the fields of en-
vironmental assessment,  engineering  and policy
analysis.  Our corporate headquarters are located in
Washington, DC, and we  maintain regional offices
in New York.  Philadelphia, Cincinnati, Chicago,
Atlanta and Dallas.

Washington Utter on Hazardous Waste
1015 18th St., NW,  Suite 200
Washington, DC 20036              202/835-2206
Washington Letter on Hazardous Waste each week
reports on  federal laws, policies, regulations and
court rulings affecting  hazardous waste  manage-
ment.  It provides  concise, independent  and in-
formed early warning to both public and private sec-
tor managers of current and coming developments
that will affect their programs and budgets.
620
            1984 EXHIBITORS

-------
Wastek, Inc.
4700 Packing House Road
Denver, CO 80216                   303/296-1765
Wastec, Inc. offers a physical-chemical waste treat-
ment system which has proven to be a commercially
viable  solution for numerous hazardous/industrial
waste  water applications. The system is compact,
with automated controls which provide intermittent
or continuous operation  upon waste demand. The
effluent from the system can be safety discharged in-
to a municipal sewage system.
 Waste Documentation Control, Inc.
 P.O. Box 7363
 Beaumont, TX 77706                409/839-4495
 Waste  Documentation  and  Control  writes  and
 markets software for documenting hazardous and
 other waste shipments. The system prints manifests,
 reports  for public agencies,  accounting reports,
 specific  format  reports and user defined reports.
 The entire system is  customized to purchaser  re-
 quirements. The system is available for many single
 or multi-terminal computers.
 Watersaver Company, Inc.
 P.O. Box 16465
 Denver, CO 80216                  303/289-1818
 Fabricators of flexible membrane liners, both rein-
 forced and unreinforced. Specializing in PVC, CPE
 and HYPALON  for  solid  waste landfills,  surface
 impoundments and hazardous waste containment.
Westbay Instruments Ltd.
507 E. Third St.
North Vancouver, B.C.
V7L1G4 Canada
604/984-4215
Westbay Instruments Ltd—Designers and manufac-
turers of the MP System, a modular multi-ported
groundwater  instrumentation system for pressure
measurements and  groundwater sampling. Com-
ponents include plastic or stainless steel casing and
couplings,  inflatable  or  mechanical  packers  and
pneumatic or electric pressure probes and sampling
probes.
Western Geophysical Corporation
P.O. Box 550
Westboro, MA 01580
617/366-9191
Weston Geophysical, for 27 years, has  provided
state-of-the-art environmental services in the United
States and abroad.  Weston's  services  include:
groundwater supply exploration  and evaluation;
hazardous waste assessment; plume evaluation; and
remedial action analysis. Weston is a leader in non-
destructive geophysical techniques with supplemen-
tal geological  capabilities. Weston offers a  newly
adapted vertical seismic profiling  technique  which
determines hydraulic conductivity of fracture zones
in  bedrock—crucial  to  contaminant  migration
analysis.

Woodward-Clyde Consultants
5120 Butler Pike
Plymouth Meeting, PA 19462       215/825-3000
Woodward-Clyde Consultants is a nationwide pro-
fessional services firm  serving clients  for over 30
years.  Our practice includes  the application of
knowledge in engineering, the earth sciences, and
the environmental and social sciences. Services of-
fered  include: waste management;  environmental
assessments; geology and hydrogeology; hydrology;
site selection studies; oil spill contingency planning;
air and water quality studies; geotechnical engineer-
ing, and risk and decision analysis.
XonTech
6862 Hayvenhurst Avenue
Van Nuys.CA 91406                818/787-7380
XonTech's  GC-810 portable,  battery  operated,
automatic gas chromatograph measures toxic emis-
sions   from   landfill  sites  with   excellent
chromatographic  separation.  It  monitors a wide
range of  halogenated  hydrocarbons of concern.
Peaks are  automatically integrated and recorded
with data on an integral printer. ECD or PID detec-
tors available. Battery  operation in field for four
hours with built-in rechargeable power pack.
                   York Laboratories
                     Div. YWC, Inc.
                   200 Monroe Turnpike
                   Monroe, CT 06468                  203/261-4458

                   Multidisciplined  environmental laboratory/engi-
                   neering consultants.  Provide air, water and waste
                   characterization  for  RCRA criteria, priority and
                   other  pollutants.  Specializing  in  gas  chroma-
                   tography/mass  spectrometry  services with  quick
                   turn-around at very competitive prices. Full service
                   environmental engineering including RCRA permit-
                   ting, hydrogeologic studies, environmental audits
                   and facility design.
                                                                                                                              1984 EXHIBITORS     621

-------