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
C! 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
------- |